This file is indexed.

/usr/share/doc/parallel/parallel.html is in parallel 20141022+ds1-1.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below.

   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
  17
  18
  19
  20
  21
  22
  23
  24
  25
  26
  27
  28
  29
  30
  31
  32
  33
  34
  35
  36
  37
  38
  39
  40
  41
  42
  43
  44
  45
  46
  47
  48
  49
  50
  51
  52
  53
  54
  55
  56
  57
  58
  59
  60
  61
  62
  63
  64
  65
  66
  67
  68
  69
  70
  71
  72
  73
  74
  75
  76
  77
  78
  79
  80
  81
  82
  83
  84
  85
  86
  87
  88
  89
  90
  91
  92
  93
  94
  95
  96
  97
  98
  99
 100
 101
 102
 103
 104
 105
 106
 107
 108
 109
 110
 111
 112
 113
 114
 115
 116
 117
 118
 119
 120
 121
 122
 123
 124
 125
 126
 127
 128
 129
 130
 131
 132
 133
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150
 151
 152
 153
 154
 155
 156
 157
 158
 159
 160
 161
 162
 163
 164
 165
 166
 167
 168
 169
 170
 171
 172
 173
 174
 175
 176
 177
 178
 179
 180
 181
 182
 183
 184
 185
 186
 187
 188
 189
 190
 191
 192
 193
 194
 195
 196
 197
 198
 199
 200
 201
 202
 203
 204
 205
 206
 207
 208
 209
 210
 211
 212
 213
 214
 215
 216
 217
 218
 219
 220
 221
 222
 223
 224
 225
 226
 227
 228
 229
 230
 231
 232
 233
 234
 235
 236
 237
 238
 239
 240
 241
 242
 243
 244
 245
 246
 247
 248
 249
 250
 251
 252
 253
 254
 255
 256
 257
 258
 259
 260
 261
 262
 263
 264
 265
 266
 267
 268
 269
 270
 271
 272
 273
 274
 275
 276
 277
 278
 279
 280
 281
 282
 283
 284
 285
 286
 287
 288
 289
 290
 291
 292
 293
 294
 295
 296
 297
 298
 299
 300
 301
 302
 303
 304
 305
 306
 307
 308
 309
 310
 311
 312
 313
 314
 315
 316
 317
 318
 319
 320
 321
 322
 323
 324
 325
 326
 327
 328
 329
 330
 331
 332
 333
 334
 335
 336
 337
 338
 339
 340
 341
 342
 343
 344
 345
 346
 347
 348
 349
 350
 351
 352
 353
 354
 355
 356
 357
 358
 359
 360
 361
 362
 363
 364
 365
 366
 367
 368
 369
 370
 371
 372
 373
 374
 375
 376
 377
 378
 379
 380
 381
 382
 383
 384
 385
 386
 387
 388
 389
 390
 391
 392
 393
 394
 395
 396
 397
 398
 399
 400
 401
 402
 403
 404
 405
 406
 407
 408
 409
 410
 411
 412
 413
 414
 415
 416
 417
 418
 419
 420
 421
 422
 423
 424
 425
 426
 427
 428
 429
 430
 431
 432
 433
 434
 435
 436
 437
 438
 439
 440
 441
 442
 443
 444
 445
 446
 447
 448
 449
 450
 451
 452
 453
 454
 455
 456
 457
 458
 459
 460
 461
 462
 463
 464
 465
 466
 467
 468
 469
 470
 471
 472
 473
 474
 475
 476
 477
 478
 479
 480
 481
 482
 483
 484
 485
 486
 487
 488
 489
 490
 491
 492
 493
 494
 495
 496
 497
 498
 499
 500
 501
 502
 503
 504
 505
 506
 507
 508
 509
 510
 511
 512
 513
 514
 515
 516
 517
 518
 519
 520
 521
 522
 523
 524
 525
 526
 527
 528
 529
 530
 531
 532
 533
 534
 535
 536
 537
 538
 539
 540
 541
 542
 543
 544
 545
 546
 547
 548
 549
 550
 551
 552
 553
 554
 555
 556
 557
 558
 559
 560
 561
 562
 563
 564
 565
 566
 567
 568
 569
 570
 571
 572
 573
 574
 575
 576
 577
 578
 579
 580
 581
 582
 583
 584
 585
 586
 587
 588
 589
 590
 591
 592
 593
 594
 595
 596
 597
 598
 599
 600
 601
 602
 603
 604
 605
 606
 607
 608
 609
 610
 611
 612
 613
 614
 615
 616
 617
 618
 619
 620
 621
 622
 623
 624
 625
 626
 627
 628
 629
 630
 631
 632
 633
 634
 635
 636
 637
 638
 639
 640
 641
 642
 643
 644
 645
 646
 647
 648
 649
 650
 651
 652
 653
 654
 655
 656
 657
 658
 659
 660
 661
 662
 663
 664
 665
 666
 667
 668
 669
 670
 671
 672
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
<?xml version="1.0" ?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title>GNU Parallel</title>
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
<link rev="made" href="mailto:root@localhost" />
</head>

<body>



<ul id="index">
  <li><a href="#NAME">NAME</a></li>
  <li><a href="#SYNOPSIS">SYNOPSIS</a></li>
  <li><a href="#DESCRIPTION">DESCRIPTION</a>
    <ul>
      <li><a href="#Readers-guide">Reader&#39;s guide</a></li>
    </ul>
  </li>
  <li><a href="#OPTIONS">OPTIONS</a></li>
  <li><a href="#EXAMPLE:-Working-as-xargs--n1.-Argument-appending">EXAMPLE: Working as xargs -n1. Argument appending</a></li>
  <li><a href="#EXAMPLE:-Reading-arguments-from-command-line">EXAMPLE: Reading arguments from command line</a></li>
  <li><a href="#EXAMPLE:-Inserting-multiple-arguments">EXAMPLE: Inserting multiple arguments</a></li>
  <li><a href="#EXAMPLE:-Context-replace">EXAMPLE: Context replace</a></li>
  <li><a href="#EXAMPLE:-Compute-intensive-jobs-and-substitution">EXAMPLE: Compute intensive jobs and substitution</a></li>
  <li><a href="#EXAMPLE:-Substitution-and-redirection">EXAMPLE: Substitution and redirection</a></li>
  <li><a href="#EXAMPLE:-Composed-commands">EXAMPLE: Composed commands</a></li>
  <li><a href="#EXAMPLE:-Calling-Bash-functions">EXAMPLE: Calling Bash functions</a></li>
  <li><a href="#EXAMPLE:-Function-tester">EXAMPLE: Function tester</a></li>
  <li><a href="#EXAMPLE:-Removing-file-extension-when-processing-files">EXAMPLE: Removing file extension when processing files</a></li>
  <li><a href="#EXAMPLE:-Removing-two-file-extensions-when-processing-files">EXAMPLE: Removing two file extensions when processing files</a></li>
  <li><a href="#EXAMPLE:-Download-10-images-for-each-of-the-past-30-days">EXAMPLE: Download 10 images for each of the past 30 days</a></li>
  <li><a href="#EXAMPLE:-Copy-files-as-last-modified-date-ISO8601-with-added-random-digits">EXAMPLE: Copy files as last modified date (ISO8601) with added random digits</a></li>
  <li><a href="#EXAMPLE:-Digtal-clock-with-blinking">EXAMPLE: Digtal clock with &quot;blinking&quot; :</a></li>
  <li><a href="#EXAMPLE:-Aggregating-content-of-files">EXAMPLE: Aggregating content of files</a></li>
  <li><a href="#EXAMPLE:-Breadth-first-parallel-web-crawler-mirrorer">EXAMPLE: Breadth first parallel web crawler/mirrorer</a></li>
  <li><a href="#EXAMPLE:-Process-files-from-a-tar-file-while-unpacking">EXAMPLE: Process files from a tar file while unpacking</a></li>
  <li><a href="#EXAMPLE:-Rewriting-a-for-loop-and-a-while-read-loop">EXAMPLE: Rewriting a for-loop and a while-read-loop</a></li>
  <li><a href="#EXAMPLE:-Rewriting-nested-for-loops">EXAMPLE: Rewriting nested for-loops</a></li>
  <li><a href="#EXAMPLE:-Finding-the-lowest-difference-between-files">EXAMPLE: Finding the lowest difference between files</a></li>
  <li><a href="#EXAMPLE:-for-loops-with-column-names">EXAMPLE: for-loops with column names</a></li>
  <li><a href="#EXAMPLE:-Count-the-differences-between-all-files-in-a-dir">EXAMPLE: Count the differences between all files in a dir</a></li>
  <li><a href="#EXAMPLE:-Speeding-up-fast-jobs">EXAMPLE: Speeding up fast jobs</a></li>
  <li><a href="#EXAMPLE:-Using-shell-variables">EXAMPLE: Using shell variables</a></li>
  <li><a href="#EXAMPLE:-Group-output-lines">EXAMPLE: Group output lines</a></li>
  <li><a href="#EXAMPLE:-Tag-output-lines">EXAMPLE: Tag output lines</a></li>
  <li><a href="#EXAMPLE:-Keep-order-of-output-same-as-order-of-input">EXAMPLE: Keep order of output same as order of input</a></li>
  <li><a href="#EXAMPLE:-Parallel-grep">EXAMPLE: Parallel grep</a></li>
  <li><a href="#EXAMPLE:-Grepping-n-lines-for-m-regular-expressions">EXAMPLE: Grepping n lines for m regular expressions.</a></li>
  <li><a href="#EXAMPLE:-Using-remote-computers">EXAMPLE: Using remote computers</a></li>
  <li><a href="#EXAMPLE:-Transferring-of-files">EXAMPLE: Transferring of files</a></li>
  <li><a href="#EXAMPLE:-Distributing-work-to-local-and-remote-computers">EXAMPLE: Distributing work to local and remote computers</a></li>
  <li><a href="#EXAMPLE:-Running-the-same-command-on-remote-computers">EXAMPLE: Running the same command on remote computers</a></li>
  <li><a href="#EXAMPLE:-Parallelizing-rsync">EXAMPLE: Parallelizing rsync</a></li>
  <li><a href="#EXAMPLE:-Use-multiple-inputs-in-one-command">EXAMPLE: Use multiple inputs in one command</a></li>
  <li><a href="#EXAMPLE:-Use-a-table-as-input">EXAMPLE: Use a table as input</a></li>
  <li><a href="#EXAMPLE:-Run-the-same-command-10-times">EXAMPLE: Run the same command 10 times</a></li>
  <li><a href="#EXAMPLE:-Working-as-cat-sh.-Resource-inexpensive-jobs-and-evaluation">EXAMPLE: Working as cat | sh. Resource inexpensive jobs and evaluation</a></li>
  <li><a href="#EXAMPLE:-Processing-a-big-file-using-more-cores">EXAMPLE: Processing a big file using more cores</a></li>
  <li><a href="#EXAMPLE:-Running-more-than-250-jobs-workaround">EXAMPLE: Running more than 250 jobs workaround</a></li>
  <li><a href="#EXAMPLE:-Working-as-mutex-and-counting-semaphore">EXAMPLE: Working as mutex and counting semaphore</a></li>
  <li><a href="#EXAMPLE:-Start-editor-with-filenames-from-stdin-standard-input">EXAMPLE: Start editor with filenames from stdin (standard input)</a></li>
  <li><a href="#EXAMPLE:-Running-sudo">EXAMPLE: Running sudo</a></li>
  <li><a href="#EXAMPLE:-GNU-Parallel-as-queue-system-batch-manager">EXAMPLE: GNU Parallel as queue system/batch manager</a></li>
  <li><a href="#EXAMPLE:-GNU-Parallel-as-dir-processor">EXAMPLE: GNU Parallel as dir processor</a></li>
  <li><a href="#QUOTING">QUOTING</a></li>
  <li><a href="#LIST-RUNNING-JOBS">LIST RUNNING JOBS</a></li>
  <li><a href="#COMPLETE-RUNNING-JOBS-BUT-DO-NOT-START-NEW-JOBS">COMPLETE RUNNING JOBS BUT DO NOT START NEW JOBS</a></li>
  <li><a href="#ENVIRONMENT-VARIABLES">ENVIRONMENT VARIABLES</a></li>
  <li><a href="#DEFAULT-PROFILE-CONFIG-FILE">DEFAULT PROFILE (CONFIG FILE)</a></li>
  <li><a href="#PROFILE-FILES">PROFILE FILES</a></li>
  <li><a href="#EXIT-STATUS">EXIT STATUS</a></li>
  <li><a href="#DIFFERENCES-BETWEEN-GNU-Parallel-AND-ALTERNATIVES">DIFFERENCES BETWEEN GNU Parallel AND ALTERNATIVES</a>
    <ul>
      <li><a href="#SUMMARY-TABLE">SUMMARY TABLE</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-xargs-AND-GNU-Parallel">DIFFERENCES BETWEEN xargs AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-find--exec-AND-GNU-Parallel">DIFFERENCES BETWEEN find -exec AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-make--j-AND-GNU-Parallel">DIFFERENCES BETWEEN make -j AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-ppss-AND-GNU-Parallel">DIFFERENCES BETWEEN ppss AND GNU Parallel</a>
        <ul>
          <li><a href="#EXAMPLES-FROM-ppss-MANUAL">EXAMPLES FROM ppss MANUAL</a></li>
        </ul>
      </li>
      <li><a href="#DIFFERENCES-BETWEEN-pexec-AND-GNU-Parallel">DIFFERENCES BETWEEN pexec AND GNU Parallel</a>
        <ul>
          <li><a href="#EXAMPLES-FROM-pexec-MANUAL">EXAMPLES FROM pexec MANUAL</a></li>
        </ul>
      </li>
      <li><a href="#DIFFERENCES-BETWEEN-xjobs-AND-GNU-Parallel">DIFFERENCES BETWEEN xjobs AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-prll-AND-GNU-Parallel">DIFFERENCES BETWEEN prll AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-dxargs-AND-GNU-Parallel">DIFFERENCES BETWEEN dxargs AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-mdm-middleman-AND-GNU-Parallel">DIFFERENCES BETWEEN mdm/middleman AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-xapply-AND-GNU-Parallel">DIFFERENCES BETWEEN xapply AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-paexec-AND-GNU-Parallel">DIFFERENCES BETWEEN paexec AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-map-AND-GNU-Parallel">DIFFERENCES BETWEEN map AND GNU Parallel</a></li>
      <li><a href="#DIFFERENCES-BETWEEN-ladon-AND-GNU-Parallel">DIFFERENCES BETWEEN ladon AND GNU Parallel</a>
        <ul>
          <li><a href="#EXAMPLES-FROM-ladon-MANUAL">EXAMPLES FROM ladon MANUAL</a></li>
        </ul>
      </li>
      <li><a href="#DIFFERENCES-BETWEEN-ClusterSSH-AND-GNU-Parallel">DIFFERENCES BETWEEN ClusterSSH AND GNU Parallel</a></li>
    </ul>
  </li>
  <li><a href="#BUGS">BUGS</a>
    <ul>
      <li><a href="#Quoting-of-newline">Quoting of newline</a></li>
      <li><a href="#Speed">Speed</a>
        <ul>
          <li><a href="#Startup">Startup</a></li>
          <li><a href="#Job-startup">Job startup</a></li>
          <li><a href="#SSH">SSH</a></li>
          <li><a href="#Disk-access">Disk access</a></li>
        </ul>
      </li>
      <li><a href="#nice-limits-command-length">--nice limits command length</a></li>
      <li><a href="#Aliases-and-functions-do-not-work">Aliases and functions do not work</a></li>
    </ul>
  </li>
  <li><a href="#REPORTING-BUGS">REPORTING BUGS</a></li>
  <li><a href="#AUTHOR">AUTHOR</a></li>
  <li><a href="#LICENSE">LICENSE</a>
    <ul>
      <li><a href="#Documentation-license-I">Documentation license I</a></li>
      <li><a href="#Documentation-license-II">Documentation license II</a></li>
    </ul>
  </li>
  <li><a href="#DEPENDENCIES">DEPENDENCIES</a></li>
  <li><a href="#SEE-ALSO">SEE ALSO</a></li>
</ul>

<h1 id="NAME">NAME</h1>

<p>parallel - build and execute shell command lines from standard input in parallel</p>

<h1 id="SYNOPSIS">SYNOPSIS</h1>

<p><b>parallel</b> [options] [<i>command</i> [arguments]] &lt; list_of_arguments</p>

<p><b>parallel</b> [options] [<i>command</i> [arguments]] ( <b>:::</b> arguments | <b>::::</b> argfile(s) ) ...</p>

<p><b>parallel</b> --semaphore [options] <i>command</i></p>

<p><b>#!/usr/bin/parallel</b> --shebang [options] [<i>command</i> [arguments]]</p>

<h1 id="DESCRIPTION">DESCRIPTION</h1>

<p>GNU <b>parallel</b> is a shell tool for executing jobs in parallel using one or more computers. A job can be a single command or a small script that has to be run for each of the lines in the input. The typical input is a list of files, a list of hosts, a list of users, a list of URLs, or a list of tables. A job can also be a command that reads from a pipe. GNU <b>parallel</b> can then split the input into blocks and pipe a block into each command in parallel.</p>

<p>If you use xargs and tee today you will find GNU <b>parallel</b> very easy to use as GNU <b>parallel</b> is written to have the same options as xargs. If you write loops in shell, you will find GNU <b>parallel</b> may be able to replace most of the loops and make them run faster by running several jobs in parallel.</p>

<p>GNU <b>parallel</b> makes sure output from the commands is the same output as you would get had you run the commands sequentially. This makes it possible to use output from GNU <b>parallel</b> as input for other programs.</p>

<p>For each line of input GNU <b>parallel</b> will execute <i>command</i> with the line as arguments. If no <i>command</i> is given, the line of input is executed. Several lines will be run in parallel. GNU <b>parallel</b> can often be used as a substitute for <b>xargs</b> or <b>cat | bash</b>.</p>

<h2 id="Readers-guide">Reader&#39;s guide</h2>

<p>Start by watching the intro videos for a quick introduction: http://www.youtube.com/playlist?list=PL284C9FF2488BC6D1</p>

<p>Then look at the <b>EXAMPLE</b>s after the list of <b>OPTIONS</b>. That will give you an idea of what GNU <b>parallel</b> is capable of.</p>

<p>Then spend an hour walking through the tutorial (<b>man parallel_tutorial</b>). Your command line will love you for it.</p>

<p>Finally you may want to look at the rest of this manual if you have special needs not already covered.</p>

<h1 id="OPTIONS">OPTIONS</h1>

<dl>

<dt id="command"><i>command</i></dt>
<dd>

<p>Command to execute. If <i>command</i> or the following arguments contain replacement strings (such as <b>{}</b>) every instance will be substituted with the input.</p>

<p>If <i>command</i> is given, GNU <b>parallel</b> solve the same tasks as <b>xargs</b>. If <i>command</i> is not given GNU <b>parallel</b> will behave similar to <b>cat | sh</b>.</p>

<p>The <i>command</i> must be an executable, a script, a composed command, or a function.</p>

<p>If it is a Bash function you need to <b>export -f</b> the function first. To use aliases copy the full environment as described under <b>--env</b> and use <b>env_parallel</b> instead of <b>parallel</b>.</p>

<p>If it is a zsh function you will need to use this helper function <b>exportf</b> to export and to set $PARALLEL_SHELL to bash:</p>

<pre><code>  function exportf (){
    export $(echo $1)=&quot;`whence -f $1 | sed -e &quot;s/$1 //&quot; `&quot;
  }

  function my_func(){
    echo $1;
    echo &quot;hello&quot;;
  }

  exportf my_func
  PARALLEL_SHELL=/bin/bash parallel &quot;my_func {}&quot; :::  1 2</code></pre>

<p>The command cannot contain the character \257 (macron: ¯).</p>

</dd>
<dt id="pod"><b>{}</b></dt>
<dd>

<p>Input line. This replacement string will be replaced by a full line read from the input source. The input source is normally stdin (standard input), but can also be given with <b>-a</b>, <b>:::</b>, or <b>::::</b>.</p>

<p>The replacement string <b>{}</b> can be changed with <b>-I</b>.</p>

<p>If the command line contains no replacement strings then <b>{}</b> will be appended to the command line.</p>

</dd>
<dt id="pod1"><b>{.}</b></dt>
<dd>

<p>Input line without extension. This replacement string will be replaced by the input with the extension removed. If the input line contains <b>.</b> after the last <b>/</b> the last <b>.</b> till the end of the string will be removed and <b>{.}</b> will be replaced with the remaining. E.g. <i>foo.jpg</i> becomes <i>foo</i>, <i>subdir/foo.jpg</i> becomes <i>subdir/foo</i>, <i>sub.dir/foo.jpg</i> becomes <i>sub.dir/foo</i>, <i>sub.dir/bar</i> remains <i>sub.dir/bar</i>. If the input line does not contain <b>.</b> it will remain unchanged.</p>

<p>The replacement string <b>{.}</b> can be changed with <b>--er</b>.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="pod2"><b>{/}</b></dt>
<dd>

<p>Basename of input line. This replacement string will be replaced by the input with the directory part removed.</p>

<p>The replacement string <b>{/}</b> can be changed with <b>--basenamereplace</b>.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="pod3"><b>{//}</b></dt>
<dd>

<p>Dirname of input line. This replacement string will be replaced by the dir of the input line. See <b>dirname</b>(1).</p>

<p>The replacement string <b>{//}</b> can be changed with <b>--dirnamereplace</b>.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="pod4"><b>{/.}</b></dt>
<dd>

<p>Basename of input line without extension. This replacement string will be replaced by the input with the directory and extension part removed. It is a combination of <b>{/}</b> and <b>{.}</b>.</p>

<p>The replacement string <b>{/.}</b> can be changed with <b>--basenameextensionreplace</b>.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="alpha-testing"><b>{#}</b> (alpha testing)</dt>
<dd>

<p>Sequence number of the job to run. This replacement string will be replaced by the sequence number of the job being run. It contains the same number as $PARALLEL_SEQ.</p>

<p>The replacement string <b>{#}</b> can be changed with <b>--seqreplace</b>.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="alpha-testing1"><b>{%}</b> (alpha testing)</dt>
<dd>

<p>Job slot number. This replacement string will be replaced by the job&#39;s slot number between 1 and number of jobs to run in parallel. There will never be 2 jobs running at the same time with the same job slot number.</p>

<p>The replacement string <b>{%}</b> can be changed with <b>--slotreplace</b>.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="n"><b>{</b><i>n</i><b>}</b></dt>
<dd>

<p>Argument from input source <i>n</i> or the <i>n</i>&#39;th argument. This positional replacement string will be replaced by the input from input source <i>n</i> (when used with <b>-a</b> or <b>::::</b>) or with the <i>n</i>&#39;th argument (when used with <b>-N</b>). If <i>n</i> is negative it refers to the <i>n</i>&#39;th last argument.</p>

<p>To understand replacement strings see <b>{}</b>.</p>

</dd>
<dt id="n1"><b>{</b><i>n</i>.<b>}</b></dt>
<dd>

<p>Argument from input source <i>n</i> or the <i>n</i>&#39;th argument without extension. It is a combination of <b>{</b><i>n</i><b>}</b> and <b>{.}</b>.</p>

<p>This positional replacement string will be replaced by the input from input source <i>n</i> (when used with <b>-a</b> or <b>::::</b>) or with the <i>n</i>&#39;th argument (when used with <b>-N</b>). The input will have the extension removed.</p>

<p>To understand positional replacement strings see <b>{</b><i>n</i><b>}</b>.</p>

</dd>
<dt id="n2"><b>{</b><i>n</i>/<b>}</b></dt>
<dd>

<p>Basename of argument from input source <i>n</i> or the <i>n</i>&#39;th argument. It is a combination of <b>{</b><i>n</i><b>}</b> and <b>{/}</b>.</p>

<p>This positional replacement string will be replaced by the input from input source <i>n</i> (when used with <b>-a</b> or <b>::::</b>) or with the <i>n</i>&#39;th argument (when used with <b>-N</b>). The input will have the directory (if any) removed.</p>

<p>To understand positional replacement strings see <b>{</b><i>n</i><b>}</b>.</p>

</dd>
<dt id="n3"><b>{</b><i>n</i>//<b>}</b></dt>
<dd>

<p>Dirname of argument from input source <i>n</i> or the <i>n</i>&#39;th argument. It is a combination of <b>{</b><i>n</i><b>}</b> and <b>{//}</b>.</p>

<p>This positional replacement string will be replaced by the dir of the input from input source <i>n</i> (when used with <b>-a</b> or <b>::::</b>) or with the <i>n</i>&#39;th argument (when used with <b>-N</b>). See <b>dirname</b>(1).</p>

<p>To understand positional replacement strings see <b>{</b><i>n</i><b>}</b>.</p>

</dd>
<dt id="n4"><b>{</b><i>n</i>/.<b>}</b></dt>
<dd>

<p>Basename of argument from input source <i>n</i> or the <i>n</i>&#39;th argument without extension. It is a combination of <b>{</b><i>n</i><b>}</b>, <b>{/}</b>, and <b>{.}</b>.</p>

<p>This positional replacement string will be replaced by the input from input source <i>n</i> (when used with <b>-a</b> or <b>::::</b>) or with the <i>n</i>&#39;th argument (when used with <b>-N</b>). The input will have the directory (if any) and extension removed.</p>

<p>To understand positional replacement strings see <b>{</b><i>n</i><b>}</b>.</p>

</dd>
<dt id="perl-expression"><b>{=</b><i>perl expression</i><b>=}</b></dt>
<dd>

<p>Replace with calculated <i>perl expression</i>. <b>$_</b> will contain the same as <b>{}</b>. After evaluating <i>perl expression</i> <b>$_</b> will be used as the value. It is recommended to only change $_ but you have full access to all of GNU <b>parallel</b>&#39;s internal functions and data structures.</p>

<p>The <b>{=</b><i>perl expression</i><b>=}</b> must be given as a single string.</p>

<p>See also: <b>--rpl</b> <b>--parens</b></p>

</dd>
<dt id="n-perl-expression"><b>{=</b><i>n</i> <i>perl expression</i><b>=}</b></dt>
<dd>

<p>Positional equivalent to <b>{= perl expression =}</b>. To understand positional replacement strings see <b>{</b><i>n</i><b>}</b>.</p>

<p>See also: <b>{= perl expression =}</b> <b>{</b><i>n</i><b>}</b>.</p>

</dd>
<dt id="arguments"><b>:::</b> <i>arguments</i></dt>
<dd>

<p>Use arguments from the command line as input source instead of stdin (standard input). Unlike other options for GNU <b>parallel</b> <b>:::</b> is placed after the <i>command</i> and before the arguments.</p>

<p>The following are equivalent:</p>

<pre><code>  (echo file1; echo file2) | parallel gzip
  parallel gzip ::: file1 file2
  parallel gzip {} ::: file1 file2
  parallel --arg-sep ,, gzip {} ,, file1 file2
  parallel --arg-sep ,, gzip ,, file1 file2
  parallel ::: &quot;gzip file1&quot; &quot;gzip file2&quot;</code></pre>

<p>To avoid treating <b>:::</b> as special use <b>--arg-sep</b> to set the argument separator to something else. See also <b>--arg-sep</b>.</p>

<p>stdin (standard input) will be passed to the first process run.</p>

<p>If multiple <b>:::</b> are given, each group will be treated as an input source, and all combinations of input sources will be generated. E.g. ::: 1 2 ::: a b c will result in the combinations (1,a) (1,b) (1,c) (2,a) (2,b) (2,c). This is useful for replacing nested for-loops.</p>

<p><b>:::</b> and <b>::::</b> can be mixed. So these are equivalent:</p>

<pre><code>  parallel echo {1} {2} {3} ::: 6 7 ::: 4 5 ::: 1 2 3
  parallel echo {1} {2} {3} :::: &lt;(seq 6 7) &lt;(seq 4 5) :::: &lt;(seq 1 3)
  parallel -a &lt;(seq 6 7) echo {1} {2} {3} :::: &lt;(seq 4 5) :::: &lt;(seq 1 3)
  parallel -a &lt;(seq 6 7) -a &lt;(seq 4 5) echo {1} {2} {3} ::: 1 2 3
  seq 6 7 | parallel -a - -a &lt;(seq 4 5) echo {1} {2} {3} ::: 1 2 3
  seq 4 5 | parallel echo {1} {2} {3} :::: &lt;(seq 6 7) - ::: 1 2 3</code></pre>

</dd>
<dt id="argfiles"><b>::::</b> <i>argfiles</i></dt>
<dd>

<p>Another way to write <b>-a</b> <i>argfile1</i> <b>-a</b> <i>argfile2</i> ...</p>

<p><b>:::</b> and <b>::::</b> can be mixed.</p>

<p>See <b>-a</b>, <b>:::</b> and <b>--xapply</b>.</p>

</dd>
<dt id="null"><b>--null</b></dt>
<dd>

</dd>
<dt id="pod-0"><b>-0</b></dt>
<dd>

<p>Use NUL as delimiter. Normally input lines will end in \n (newline). If they end in \0 (NUL), then use this option. It is useful for processing arguments that may contain \n (newline).</p>

</dd>
<dt id="arg-file-input-file"><b>--arg-file</b> <i>input-file</i></dt>
<dd>

</dd>
<dt id="a-input-file"><b>-a</b> <i>input-file</i></dt>
<dd>

<p>Use <i>input-file</i> as input source. If you use this option, stdin (standard input) is given to the first process run. Otherwise, stdin (standard input) is redirected from /dev/null.</p>

<p>If multiple <b>-a</b> are given, each <i>input-file</i> will be treated as an input source, and all combinations of input sources will be generated. E.g. The file <b>foo</b> contains <b>1 2</b>, the file <b>bar</b> contains <b>a b c</b>. <b>-a foo</b> <b>-a bar</b> will result in the combinations (1,a) (1,b) (1,c) (2,a) (2,b) (2,c). This is useful for replacing nested for-loops.</p>

<p>See also <b>--xapply</b> and <b>{</b><i>n</i><b>}</b>.</p>

</dd>
<dt id="arg-file-sep-sep-str"><b>--arg-file-sep</b> <i>sep-str</i></dt>
<dd>

<p>Use <i>sep-str</i> instead of <b>::::</b> as separator string between command and argument files. Useful if <b>::::</b> is used for something else by the command.</p>

<p>See also: <b>::::</b>.</p>

</dd>
<dt id="arg-sep-sep-str"><b>--arg-sep</b> <i>sep-str</i></dt>
<dd>

<p>Use <i>sep-str</i> instead of <b>:::</b> as separator string. Useful if <b>:::</b> is used for something else by the command.</p>

<p>Also useful if you command uses <b>:::</b> but you still want to read arguments from stdin (standard input): Simply change <b>--arg-sep</b> to a string that is not in the command line.</p>

<p>See also: <b>:::</b>.</p>

</dd>
<dt id="bar"><b>--bar</b></dt>
<dd>

<p>Show progress as a progress bar. In the bar is shown: % of jobs completed, estimated seconds left, and number of jobs started.</p>

<p>It is compatible with <b>zenity</b>:</p>

<p>seq 1000 | parallel -j30 --bar &#39;(echo {};sleep 0.1)&#39; 2&gt; &gt;(zenity --progress --auto-kill) | wc</p>

</dd>
<dt id="basefile-file"><b>--basefile</b> <i>file</i></dt>
<dd>

</dd>
<dt id="bf-file"><b>--bf</b> <i>file</i></dt>
<dd>

<p><i>file</i> will be transferred to each sshlogin before a jobs is started. It will be removed if <b>--cleanup</b> is active. The file may be a script to run or some common base data needed for the jobs. Multiple <b>--bf</b> can be specified to transfer more basefiles. The <i>file</i> will be transferred the same way as <b>--transfer</b>.</p>

</dd>
<dt id="basenamereplace-replace-str"><b>--basenamereplace</b> <i>replace-str</i></dt>
<dd>

</dd>
<dt id="bnr-replace-str"><b>--bnr</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of <b>{/}</b> for basename of input line.</p>

</dd>
<dt id="basenameextensionreplace-replace-str"><b>--basenameextensionreplace</b> <i>replace-str</i></dt>
<dd>

</dd>
<dt id="bner-replace-str"><b>--bner</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of <b>{/.}</b> for basename of input line without extension.</p>

</dd>
<dt id="bg"><b>--bg</b></dt>
<dd>

<p>Run command in background thus GNU <b>parallel</b> will not wait for completion of the command before exiting. This is the default if <b>--semaphore</b> is set.</p>

<p>See also: <b>--fg</b>, <b>man sem</b>.</p>

<p>Implies <b>--semaphore</b>.</p>

</dd>
<dt id="bibtex"><b>--bibtex</b></dt>
<dd>

<p>Print the BibTeX entry for GNU <b>parallel</b> and disable citation notice.</p>

</dd>
<dt id="block-size"><b>--block</b> <i>size</i></dt>
<dd>

</dd>
<dt id="block-size-size"><b>--block-size</b> <i>size</i></dt>
<dd>

<p>Size of block in bytes. The size can be postfixed with K, M, G, T, P, k, m, g, t, or p which would multiply the size with 1024, 1048576, 1073741824, 1099511627776, 1125899906842624, 1000, 1000000, 1000000000, 1000000000000, or 1000000000000000 respectively.</p>

<p>GNU <b>parallel</b> tries to meet the block size but can be off by the length of one record. For performance reasons <i>size</i> should be bigger than a single record.</p>

<p><i>size</i> defaults to 1M.</p>

<p>See <b>--pipe</b> for use of this.</p>

</dd>
<dt id="cat"><b>--cat</b></dt>
<dd>

<p>Create a temporary file with content. Normally <b>--pipe</b> will give data to the program on stdin (standard input). With <b>--cat</b> GNU <b>parallel</b> will create a temporary file with the name in {}, so you can do: <b>parallel --pipe --cat wc {}</b>.</p>

<p>See also <b>--fifo</b>.</p>

</dd>
<dt id="cleanup"><b>--cleanup</b></dt>
<dd>

<p>Remove transferred files. <b>--cleanup</b> will remove the transferred files on the remote computer after processing is done.</p>

<pre><code>  find log -name &#39;*gz&#39; | parallel \
    --sshlogin server.example.com --transfer --return {.}.bz2 \
    --cleanup &quot;zcat {} | bzip -9 &gt;{.}.bz2&quot;</code></pre>

<p>With <b>--transfer</b> the file transferred to the remote computer will be removed on the remote computer. Directories created will not be removed - even if they are empty.</p>

<p>With <b>--return</b> the file transferred from the remote computer will be removed on the remote computer. Directories created will not be removed - even if they are empty.</p>

<p><b>--cleanup</b> is ignored when not used with <b>--transfer</b> or <b>--return</b>.</p>

</dd>
<dt id="colsep-regexp"><b>--colsep</b> <i>regexp</i></dt>
<dd>

</dd>
<dt id="C-regexp"><b>-C</b> <i>regexp</i></dt>
<dd>

<p>Column separator. The input will be treated as a table with <i>regexp</i> separating the columns. The n&#39;th column can be access using <b>{</b><i>n</i><b>}</b> or <b>{</b><i>n</i>.<b>}</b>. E.g. <b>{3}</b> is the 3rd column.</p>

<p><b>--colsep</b> implies <b>--trim rl</b>.</p>

<p><i>regexp</i> is a Perl Regular Expression: http://perldoc.perl.org/perlre.html</p>

</dd>
<dt id="compress"><b>--compress</b></dt>
<dd>

<p>Compress temporary files. If the output is big and very compressible this will take up less disk space in $TMPDIR and possibly be faster due to less disk I/O.</p>

<p>GNU <b>parallel</b> will try <b>lzop</b>, <b>pigz</b>, <b>gzip</b>, <b>pbzip2</b>, <b>plzip</b>, <b>bzip2</b>, <b>lzma</b>, <b>lzip</b>, <b>xz</b> in that order, and use the first available.</p>

</dd>
<dt id="compress-program-prg"><b>--compress-program</b> <i>prg</i></dt>
<dd>

</dd>
<dt id="decompress-program-prg"><b>--decompress-program</b> <i>prg</i></dt>
<dd>

<p>Use <i>prg</i> for (de)compressing temporary files. It is assumed that <i>prg -dc</i> will decompress stdin (standard input) to stdout (standard output) unless <b>--decompress-program</b> is given.</p>

</dd>
<dt id="ctrlc"><b>--ctrlc</b></dt>
<dd>

<p>Sends SIGINT to tasks running on remote computers thus killing them.</p>

</dd>
<dt id="delimiter-delim"><b>--delimiter</b> <i>delim</i></dt>
<dd>

</dd>
<dt id="d-delim"><b>-d</b> <i>delim</i></dt>
<dd>

<p>Input items are terminated by the specified character. Quotes and backslash are not special; every character in the input is taken literally. Disables the end-of-file string, which is treated like any other argument. This can be used when the input consists of simply newline-separated items, although it is almost always better to design your program to use --null where this is possible. The specified delimiter may be a single character, a C-style character escape such as \n, or an octal or hexadecimal escape code. Octal and hexadecimal escape codes are understood as for the printf command. Multibyte characters are not supported.</p>

</dd>
<dt id="dirnamereplace-replace-str"><b>--dirnamereplace</b> <i>replace-str</i></dt>
<dd>

</dd>
<dt id="dnr-replace-str"><b>--dnr</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of <b>{//}</b> for dirname of input line.</p>

</dd>
<dt id="E-eof-str"><b>-E</b> <i>eof-str</i></dt>
<dd>

<p>Set the end of file string to eof-str. If the end of file string occurs as a line of input, the rest of the input is ignored. If neither <b>-E</b> nor <b>-e</b> is used, no end of file string is used.</p>

</dd>
<dt id="delay-secs"><b>--delay</b> <i>secs</i></dt>
<dd>

<p>Delay starting next job <i>secs</i> seconds. GNU <b>parallel</b> will pause <i>secs</i> seconds after starting each job. <i>secs</i> can be less than 1 seconds.</p>

</dd>
<dt id="dry-run"><b>--dry-run</b></dt>
<dd>

<p>Print the job to run on stdout (standard output), but do not run the job. Use <b>-v -v</b> to include the ssh/rsync wrapping if the job would be run on a remote computer. Do not count on this literaly, though, as the job may be scheduled on another computer or the local computer if : is in the list.</p>

</dd>
<dt id="eof-eof-str"><b>--eof</b>[=<i>eof-str</i>]</dt>
<dd>

</dd>
<dt id="e-eof-str"><b>-e</b>[<i>eof-str</i>]</dt>
<dd>

<p>This option is a synonym for the <b>-E</b> option. Use <b>-E</b> instead, because it is POSIX compliant for <b>xargs</b> while this option is not. If <i>eof-str</i> is omitted, there is no end of file string. If neither <b>-E</b> nor <b>-e</b> is used, no end of file string is used.</p>

</dd>
<dt id="env-var-alpha-testing"><b>--env</b> <i>var</i> (alpha testing)</dt>
<dd>

<p>Copy environment variable <i>var</i>. This will copy <i>var</i> to the environment that the command is run in. This is especially useful for remote execution.</p>

<p>In Bash <i>var</i> can also be a Bash function - just remember to <b>export -f</b> the function, see <b>command</b>.</p>

<p>The variable &#39;_&#39; is special. It will copy all enviroment variables except for the ones mentioned in ~/.parallel/ignored_vars.</p>

<p>To copy Bash arrays you need an importer function, as Bash arrays cannot be exported:</p>

<pre><code>    import_array () {
      local func=$1; shift;
      export $func=&#39;() {
        &#39;&quot;$(for arr in $@; do
              declare -p $arr|sed &#39;1s/declare -./&amp;g/&#39;
            done)&quot;&#39;
      }&#39;
    }

    declare -A assoc=&#39;([one]=&quot;1&quot; [two]=&quot;2&quot;)&#39;
    declare -a indexed=&#39;([0]=&quot;one&quot; [1]=&quot;two&quot;)&#39;

    import_array my_importer assoc indexed

    parallel --env my_importer \
      &#39;my_importer; echo &quot;{}&quot; &quot;${indexed[{}]}&quot; &quot;${assoc[${indexed[{}]}]}&quot;&#39; ::: &quot;${!indexed[@]}&quot;</code></pre>

<p>To copy the full environment use this function (e.g. by putting it in .bashrc):</p>

<pre><code>    env_parallel() {
      export parallel_bash_environment=&#39;() {
        &#39;&quot;$(echo &quot;shopt -s expand_aliases 2&gt;/dev/null&quot;; alias;typeset -p | grep -vFf &lt;(readonly; echo GROUPS; echo FUNCNAME; echo DIRSTACK; echo _; echo PIPESTATUS; echo USERNAME) | grep -v BASH_;typeset -f)&quot;&#39;
      }&#39;
      # Run as: env_parallel [normal parallel options]
      `which parallel` &quot;$@&quot;
      unset parallel_bash_environment
    }
    # call as:
    env_parallel [normal parallel options]</code></pre>

<p>See also: <b>--record-env</b>.</p>

</dd>
<dt id="eta"><b>--eta</b></dt>
<dd>

<p>Show the estimated number of seconds before finishing. This forces GNU <b>parallel</b> to read all jobs before starting to find the number of jobs. GNU <b>parallel</b> normally only reads the next job to run. Implies <b>--progress</b>.</p>

</dd>
<dt id="fg"><b>--fg</b></dt>
<dd>

<p>Run command in foreground thus GNU <b>parallel</b> will wait for completion of the command before exiting.</p>

<p>See also <b>--bg</b>, <b>man sem</b>.</p>

<p>Implies <b>--semaphore</b>.</p>

</dd>
<dt id="fifo"><b>--fifo</b></dt>
<dd>

<p>Create a temporary fifo with content. Normally <b>--pipe</b> will give data to the program on stdin (standard input). With <b>--fifo</b> GNU <b>parallel</b> will create a temporary fifo with the name in {}, so you can do: <b>parallel --pipe --fifo wc {}</b>.</p>

<p>Beware: If data is not read from the fifo, the job will block forever.</p>

<p>See also <b>--cat</b>.</p>

</dd>
<dt id="filter-hosts"><b>--filter-hosts</b></dt>
<dd>

<p>Remove down hosts. For each remote host: check that login through ssh works. If not: do not use this host.</p>

<p>Currently you can <i>not</i> put <b>--filter-hosts</b> in a profile, $PARALLEL, /etc/parallel/config or similar. This is because GNU <b>parallel</b> uses GNU <b>parallel</b> to compute this, so you will get an infinite loop. This will likely be fixed in a later release.</p>

</dd>
<dt id="gnu"><b>--gnu</b></dt>
<dd>

<p>Behave like GNU <b>parallel</b>. This option historically took precedence over <b>--tollef</b>. The <b>--tollef</b> option is now retired, and therefore may not be used. <b>--gnu</b> is kept for compatibility.</p>

</dd>
<dt id="group"><b>--group</b></dt>
<dd>

<p>Group output. Output from each jobs is grouped together and is only printed when the command is finished. stderr (standard error) first followed by stdout (standard output). This takes some CPU time. In rare situations GNU <b>parallel</b> takes up lots of CPU time and if it is acceptable that the outputs from different commands are mixed together, then disabling grouping with <b>-u</b> can speedup GNU <b>parallel</b> by a factor of 10.</p>

<p><b>--group</b> is the default. Can be reversed with <b>-u</b>.</p>

<p>See also: <b>--line-buffer</b> <b>--ungroup</b></p>

</dd>
<dt id="help"><b>--help</b></dt>
<dd>

</dd>
<dt id="h"><b>-h</b></dt>
<dd>

<p>Print a summary of the options to GNU <b>parallel</b> and exit.</p>

</dd>
<dt id="halt-on-error-val"><b>--halt-on-error</b> <i>val</i></dt>
<dd>

</dd>
<dt id="halt-val"><b>--halt</b> <i>val</i></dt>
<dd>

<p>How should GNU <b>parallel</b> terminate if one of more jobs fail?</p>

<dl>

<dt id="pod0">0</dt>
<dd>

<p>Do not halt if a job fails. Exit status will be the number of jobs failed. This is the default.</p>

</dd>
<dt id="pod11">1</dt>
<dd>

<p>Do not start new jobs if a job fails, but complete the running jobs including cleanup. The exit status will be the exit status from the last failing job.</p>

</dd>
<dt id="pod21">2</dt>
<dd>

<p>Kill off all jobs immediately and exit without cleanup. The exit status will be the exit status from the failing job.</p>

</dd>
<dt id="pod1-99">1-99%</dt>
<dd>

<p>If <i>val</i>% of the jobs fail and minimum 3: Do not start new jobs, but complete the running jobs including cleanup. The exit status will be the exit status from the last failing job.</p>

</dd>
</dl>

</dd>
<dt id="header-regexp"><b>--header</b> <i>regexp</i></dt>
<dd>

<p>Use regexp as header. For normal usage the matched header (typically the first line: <b>--header &#39;.*\n&#39;</b>) will be split using <b>--colsep</b> (which will default to &#39;\t&#39;) and column names can be used as replacement variables: <b>{column name}</b>.</p>

<p>For <b>--pipe</b> the matched header will be prepended to each output.</p>

<p><b>--header :</b> is an alias for <b>--header &#39;.*\n&#39;</b>.</p>

<p>If <i>regexp</i> is a number, it will match that many lines.</p>

</dd>
<dt id="I-replace-str"><b>-I</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of {}.</p>

</dd>
<dt id="replace-replace-str"><b>--replace</b>[=<i>replace-str</i>]</dt>
<dd>

</dd>
<dt id="i-replace-str"><b>-i</b>[<i>replace-str</i>]</dt>
<dd>

<p>This option is a synonym for <b>-I</b><i>replace-str</i> if <i>replace-str</i> is specified, and for <b>-I</b>{} otherwise. This option is deprecated; use <b>-I</b> instead.</p>

</dd>
<dt id="joblog-logfile"><b>--joblog</b> <i>logfile</i></dt>
<dd>

<p>Logfile for executed jobs. Save a list of the executed jobs to <i>logfile</i> in the following TAB separated format: sequence number, sshlogin, start time as seconds since epoch, run time in seconds, bytes in files transferred, bytes in files returned, exit status, signal, and command run.</p>

<p>To convert the times into ISO-8601 strict do:</p>

<p><b>perl -a -F&quot;\t&quot; -ne &#39;chomp($F[2]=`date -d \@$F[2] +%FT%T`); print join(&quot;\t&quot;,@F)&#39;</b></p>

<p>See also <b>--resume</b>.</p>

</dd>
<dt id="jobs-N"><b>--jobs</b> <i>N</i></dt>
<dd>

</dd>
<dt id="j-N"><b>-j</b> <i>N</i></dt>
<dd>

</dd>
<dt id="max-procs-N"><b>--max-procs</b> <i>N</i></dt>
<dd>

</dd>
<dt id="P-N"><b>-P</b> <i>N</i></dt>
<dd>

<p>Number of jobslots. Run up to N jobs in parallel. 0 means as many as possible. Default is 100% which will run one job per CPU core.</p>

<p>If <b>--semaphore</b> is set default is 1 thus making a mutex.</p>

</dd>
<dt id="jobs-N1"><b>--jobs</b> <i>+N</i></dt>
<dd>

</dd>
<dt id="j-N1"><b>-j</b> <i>+N</i></dt>
<dd>

</dd>
<dt id="max-procs-N1"><b>--max-procs</b> <i>+N</i></dt>
<dd>

</dd>
<dt id="P-N1"><b>-P</b> <i>+N</i></dt>
<dd>

<p>Add N to the number of CPU cores. Run this many jobs in parallel. See also <b>--use-cpus-instead-of-cores</b>.</p>

</dd>
<dt id="jobs--N"><b>--jobs</b> <i>-N</i></dt>
<dd>

</dd>
<dt id="j--N"><b>-j</b> <i>-N</i></dt>
<dd>

</dd>
<dt id="max-procs--N"><b>--max-procs</b> <i>-N</i></dt>
<dd>

</dd>
<dt id="P--N"><b>-P</b> <i>-N</i></dt>
<dd>

<p>Subtract N from the number of CPU cores. Run this many jobs in parallel. If the evaluated number is less than 1 then 1 will be used. See also <b>--use-cpus-instead-of-cores</b>.</p>

</dd>
<dt id="jobs-N2"><b>--jobs</b> <i>N</i>%</dt>
<dd>

</dd>
<dt id="j-N2"><b>-j</b> <i>N</i>%</dt>
<dd>

</dd>
<dt id="max-procs-N2"><b>--max-procs</b> <i>N</i>%</dt>
<dd>

</dd>
<dt id="P-N2"><b>-P</b> <i>N</i>%</dt>
<dd>

<p>Multiply N% with the number of CPU cores. Run this many jobs in parallel. See also <b>--use-cpus-instead-of-cores</b>.</p>

</dd>
<dt id="jobs-procfile"><b>--jobs</b> <i>procfile</i></dt>
<dd>

</dd>
<dt id="j-procfile"><b>-j</b> <i>procfile</i></dt>
<dd>

</dd>
<dt id="max-procs-procfile"><b>--max-procs</b> <i>procfile</i></dt>
<dd>

</dd>
<dt id="P-procfile"><b>-P</b> <i>procfile</i></dt>
<dd>

<p>Read parameter from file. Use the content of <i>procfile</i> as parameter for <i>-j</i>. E.g. <i>procfile</i> could contain the string 100% or +2 or 10. If <i>procfile</i> is changed when a job completes, <i>procfile</i> is read again and the new number of jobs is computed. If the number is lower than before, running jobs will be allowed to finish but new jobs will not be started until the wanted number of jobs has been reached. This makes it possible to change the number of simultaneous running jobs while GNU <b>parallel</b> is running.</p>

</dd>
<dt id="keep-order"><b>--keep-order</b></dt>
<dd>

</dd>
<dt id="k"><b>-k</b></dt>
<dd>

<p>Keep sequence of output same as the order of input. Normally the output of a job will be printed as soon as the job completes. Try this to see the difference:</p>

<pre><code>  parallel -j4 sleep {}\; echo {} ::: 2 1 4 3
  parallel -j4 -k sleep {}\; echo {} ::: 2 1 4 3</code></pre>

<p>If used with <b>--onall</b> or <b>--nonall</b> the output will grouped by sshlogin in sorted order.</p>

</dd>
<dt id="L-max-lines"><b>-L</b> <i>max-lines</i></dt>
<dd>

<p>When used with <b>--pipe</b>: Read records of <i>max-lines</i>.</p>

<p>When used otherwise: Use at most <i>max-lines</i> nonblank input lines per command line. Trailing blanks cause an input line to be logically continued on the next input line.</p>

<p><b>-L 0</b> means read one line, but insert 0 arguments on the command line.</p>

<p>Implies <b>-X</b> unless <b>-m</b>, <b>--xargs</b>, or <b>--pipe</b> is set.</p>

</dd>
<dt id="max-lines-max-lines"><b>--max-lines</b>[=<i>max-lines</i>]</dt>
<dd>

</dd>
<dt id="l-max-lines"><b>-l</b>[<i>max-lines</i>]</dt>
<dd>

<p>When used with <b>--pipe</b>: Read records of <i>max-lines</i>.</p>

<p>When used otherwise: Synonym for the <b>-L</b> option. Unlike <b>-L</b>, the <i>max-lines</i> argument is optional. If <i>max-lines</i> is not specified, it defaults to one. The <b>-l</b> option is deprecated since the POSIX standard specifies <b>-L</b> instead.</p>

<p><b>-l 0</b> is an alias for <b>-l 1</b>.</p>

<p>Implies <b>-X</b> unless <b>-m</b>, <b>--xargs</b>, or <b>--pipe</b> is set.</p>

</dd>
<dt id="line-buffer"><b>--line-buffer</b></dt>
<dd>

<p>Buffer output on line basis. <b>--group</b> will keep the output together for a whole job. <b>--ungroup</b> allows output to mixup with half a line coming from one job and half a line coming from another job. <b>--line-buffer</b> fits between these two: GNU <b>parallel</b> will print a full line, but will allow for mixing lines of different jobs.</p>

<p><b>--line-buffer</b> takes more CPU power than than both <b>--group</b> and <b>--ungroup</b>, but can be faster than <b>--group</b> if the CPU is not the limiting factor.</p>

<p>See also: <b>--group</b> <b>--ungroup</b></p>

</dd>
<dt id="load-max-load"><b>--load</b> <i>max-load</i></dt>
<dd>

<p>Do not start new jobs on a given computer unless the number of running processes on the computer is less than <i>max-load</i>. <i>max-load</i> uses the same syntax as <b>--jobs</b>, so <i>100%</i> for one per CPU is a valid setting. Only difference is 0 which is interpreted as 0.01.</p>

</dd>
<dt id="controlmaster"><b>--controlmaster</b></dt>
<dd>

</dd>
<dt id="M"><b>-M</b></dt>
<dd>

<p>Use ssh&#39;s ControlMaster to make ssh connections faster. Useful if jobs run remote and are very fast to run. This is disabled for sshlogins that specify their own ssh command.</p>

</dd>
<dt id="xargs"><b>--xargs</b></dt>
<dd>

<p>Multiple arguments. Insert as many arguments as the command line length permits.</p>

<p>If <b>{}</b> is not used the arguments will be appended to the line. If <b>{}</b> is used multiple times each <b>{}</b> will be replaced with all the arguments.</p>

<p>Support for <b>--xargs</b> with <b>--sshlogin</b> is limited and may fail.</p>

<p>See also <b>-X</b> for context replace. If in doubt use <b>-X</b> as that will most likely do what is needed.</p>

</dd>
<dt id="m"><b>-m</b></dt>
<dd>

<p>Multiple arguments. Insert as many arguments as the command line length permits. If multiple jobs are being run in parallel: distribute the arguments evenly among the jobs. Use <b>-j1</b> to avoid this.</p>

<p>If <b>{}</b> is not used the arguments will be appended to the line. If <b>{}</b> is used multiple times each <b>{}</b> will be replaced with all the arguments.</p>

<p>Support for <b>-m</b> with <b>--sshlogin</b> is limited and may fail.</p>

<p>See also <b>-X</b> for context replace. If in doubt use <b>-X</b> as that will most likely do what is needed.</p>

</dd>
<dt id="minversion-version"><b>--minversion</b> <i>version</i></dt>
<dd>

<p>Print the version GNU <b>parallel</b> and exit. If the current version of GNU <b>parallel</b> is less than <i>version</i> the exit code is 255. Otherwise it is 0.</p>

<p>This is useful for scripts that depend on features only available from a certain version of GNU <b>parallel</b>.</p>

</dd>
<dt id="nonall-beta-testing"><b>--nonall</b> (beta testing)</dt>
<dd>

<p><b>--onall</b> with no arguments. Run the command on all computers given with <b>--sshlogin</b> but take no arguments. GNU <b>parallel</b> will log into <b>--jobs</b> number of computers in parallel and run the job on the computer. <b>-j</b> adjusts how many computers to log into in parallel.</p>

<p>This is useful for running the same command (e.g. uptime) on a list of servers.</p>

</dd>
<dt id="onall-beta-testing"><b>--onall</b> (beta testing)</dt>
<dd>

<p>Run all the jobs on all computers given with <b>--sshlogin</b>. GNU <b>parallel</b> will log into <b>--jobs</b> number of computers in parallel and run one job at a time on the computer. The order of the jobs will not be changed, but some computers may finish before others. <b>-j</b> adjusts how many computers to log into in parallel.</p>

<p>When using <b>--group</b> the output will be grouped by each server, so all the output from one server will be grouped together.</p>

</dd>
<dt id="output-as-files"><b>--output-as-files</b></dt>
<dd>

</dd>
<dt id="outputasfiles"><b>--outputasfiles</b></dt>
<dd>

</dd>
<dt id="files"><b>--files</b></dt>
<dd>

<p>Instead of printing the output to stdout (standard output) the output of each job is saved in a file and the filename is then printed.</p>

</dd>
<dt id="pipe"><b>--pipe</b></dt>
<dd>

</dd>
<dt id="spreadstdin"><b>--spreadstdin</b></dt>
<dd>

<p>Spread input to jobs on stdin (standard input). Read a block of data from stdin (standard input) and give one block of data as input to one job.</p>

<p>The block size is determined by <b>--block</b>. The strings <b>--recstart</b> and <b>--recend</b> tell GNU <b>parallel</b> how a record starts and/or ends. The block read will have the final partial record removed before the block is passed on to the job. The partial record will be prepended to next block.</p>

<p>If <b>--recstart</b> is given this will be used to split at record start.</p>

<p>If <b>--recend</b> is given this will be used to split at record end.</p>

<p>If both <b>--recstart</b> and <b>--recend</b> are given both will have to match to find a split position.</p>

<p>If neither <b>--recstart</b> nor <b>--recend</b> are given <b>--recend</b> defaults to &#39;\n&#39;. To have no record separator use <b>--recend &quot;&quot;</b>.</p>

<p><b>--files</b> is often used with <b>--pipe</b>.</p>

<p>See also: <b>--recstart</b>, <b>--recend</b>, <b>--fifo</b>, <b>--cat</b>, <b>--pipepart</b>.</p>

</dd>
<dt id="pipepart-beta-testing"><b>--pipepart</b> (beta testing)</dt>
<dd>

<p>Pipe parts of a physical file. <b>--pipepart</b> works similar to <b>--pipe</b>, but is much faster. It has a few limitations:</p>

<dl>

<dt id="pod5">*</dt>
<dd>

<p>The file must be a physical (seekable) file and must be given using <b>-a</b> or <b>::::</b>.</p>

</dd>
<dt id="pod6">*</dt>
<dd>

<p>Record counting (<b>-N</b>) and line counting (<b>-L</b>/<b>-l</b>) do not work.</p>

</dd>
</dl>

</dd>
<dt id="plain"><b>--plain</b></dt>
<dd>

<p>Ignore any <b>--profile</b>, $PARALLEL, and ~/.parallel/config to get full control on the command line (used by GNU <b>parallel</b> internally when called with <b>--sshlogin</b>).</p>

</dd>
<dt id="plus"><b>--plus</b></dt>
<dd>

<p>Activate additional replacement strings: {+/} {+.} {+..} {+...} {..} {...} {/..} {/...}. The idea being that &#39;{+foo}&#39; matches the opposite of &#39;{foo}&#39; and {} = {+/}/{/} = {.}.{+.} = {+/}/{/.}.{+.} = {..}.{+..} = {+/}/{/..}.{+..} = {...}.{+...} = {+/}/{/...}.{+...}</p>

</dd>
<dt id="progress"><b>--progress</b></dt>
<dd>

<p>Show progress of computations. List the computers involved in the task with number of CPU cores detected and the max number of jobs to run. After that show progress for each computer: number of running jobs, number of completed jobs, and percentage of all jobs done by this computer. The percentage will only be available after all jobs have been scheduled as GNU <b>parallel</b> only read the next job when ready to schedule it - this is to avoid wasting time and memory by reading everything at startup.</p>

<p>By sending GNU <b>parallel</b> SIGUSR2 you can toggle turning on/off <b>--progress</b> on a running GNU <b>parallel</b> process.</p>

<p>See also <b>--eta</b>.</p>

</dd>
<dt id="max-args-max-args"><b>--max-args</b>=<i>max-args</i></dt>
<dd>

</dd>
<dt id="n-max-args"><b>-n</b> <i>max-args</i></dt>
<dd>

<p>Use at most <i>max-args</i> arguments per command line. Fewer than <i>max-args</i> arguments will be used if the size (see the <b>-s</b> option) is exceeded, unless the <b>-x</b> option is given, in which case GNU <b>parallel</b> will exit.</p>

<p><b>-n 0</b> means read one argument, but insert 0 arguments on the command line.</p>

<p>Implies <b>-X</b> unless <b>-m</b> is set.</p>

</dd>
<dt id="max-replace-args-max-args"><b>--max-replace-args</b>=<i>max-args</i></dt>
<dd>

</dd>
<dt id="N-max-args"><b>-N</b> <i>max-args</i></dt>
<dd>

<p>Use at most <i>max-args</i> arguments per command line. Like <b>-n</b> but also makes replacement strings <b>{1}</b> .. <b>{</b><i>max-args</i><b>}</b> that represents argument 1 .. <i>max-args</i>. If too few args the <b>{</b><i>n</i><b>}</b> will be empty.</p>

<p><b>-N 0</b> means read one argument, but insert 0 arguments on the command line.</p>

<p>This will set the owner of the homedir to the user:</p>

<p><b>tr &#39;:&#39; &#39;\n&#39; &lt; /etc/passwd | parallel -N7 chown {1} {6}</b></p>

<p>Implies <b>-X</b> unless <b>-m</b> or <b>--pipe</b> is set.</p>

<p>When used with <b>--pipe</b> <b>-N</b> is the number of records to read. This is somewhat slower than <b>--block</b>.</p>

</dd>
<dt id="max-line-length-allowed"><b>--max-line-length-allowed</b></dt>
<dd>

<p>Print the maximal number of characters allowed on the command line and exit (used by GNU <b>parallel</b> itself to determine the line length on remote computers).</p>

</dd>
<dt id="number-of-cpus"><b>--number-of-cpus</b></dt>
<dd>

<p>Print the number of physical CPUs and exit (used by GNU <b>parallel</b> itself to determine the number of physical CPUs on remote computers).</p>

</dd>
<dt id="number-of-cores"><b>--number-of-cores</b></dt>
<dd>

<p>Print the number of CPU cores and exit (used by GNU <b>parallel</b> itself to determine the number of CPU cores on remote computers).</p>

</dd>
<dt id="no-notice"><b>--no-notice</b></dt>
<dd>

<p>Do not display citation notice. A citation notice is printed on stderr (standard error) only if stderr (standard error) is a terminal, the user has not specified <b>--no-notice</b>, and the user has not run <b>--bibtex</b> once.</p>

</dd>
<dt id="nice-niceness"><b>--nice</b> <i>niceness</i></dt>
<dd>

<p>Run the command at this niceness. For simple commands you can just add <b>nice</b> in front of the command. But if the command consists of more sub commands (Like: ls|wc) then prepending <b>nice</b> will not always work. <b>--nice</b> will make sure all sub commands are niced.</p>

</dd>
<dt id="interactive"><b>--interactive</b></dt>
<dd>

</dd>
<dt id="p"><b>-p</b></dt>
<dd>

<p>Prompt the user about whether to run each command line and read a line from the terminal. Only run the command line if the response starts with &#39;y&#39; or &#39;Y&#39;. Implies <b>-t</b>.</p>

</dd>
<dt id="parens-parensstring"><b>--parens</b> <i>parensstring</i></dt>
<dd>

<p>Use to define start and end parenthesis for <b>{= perl expression =}</b>. The left and the right parenthesis can be multiple characters and are assumed to be the same length. The default is <b>{==}</b> giving <b>{=</b> as the start parenthesis and <b>=}</b> as the end parenthesis.</p>

<p>Another useful setting is <b>,,,,</b> which would make both parenthesis <b>,,</b>:</p>

<pre><code>  parallel --parens ,,,, echo foo is ,,s/I/O/g,, ::: FII</code></pre>

<p>See also: <b>--rpl</b> <b>{= perl expression =}</b></p>

</dd>
<dt id="profile-profilename"><b>--profile</b> <i>profilename</i></dt>
<dd>

</dd>
<dt id="J-profilename"><b>-J</b> <i>profilename</i></dt>
<dd>

<p>Use profile <i>profilename</i> for options. This is useful if you want to have multiple profiles. You could have one profile for running jobs in parallel on the local computer and a different profile for running jobs on remote computers. See the section PROFILE FILES for examples.</p>

<p><i>profilename</i> corresponds to the file ~/.parallel/<i>profilename</i>.</p>

<p>You can give multiple profiles by repeating <b>--profile</b>. If parts of the profiles conflict, the later ones will be used.</p>

<p>Default: config</p>

</dd>
<dt id="quote"><b>--quote</b></dt>
<dd>

</dd>
<dt id="q"><b>-q</b></dt>
<dd>

<p>Quote <i>command</i>. This will quote the command line so special characters are not interpreted by the shell. See the section QUOTING. Most people will never need this. Quoting is disabled by default.</p>

</dd>
<dt id="no-run-if-empty"><b>--no-run-if-empty</b></dt>
<dd>

</dd>
<dt id="r"><b>-r</b></dt>
<dd>

<p>If the stdin (standard input) only contains whitespace, do not run the command.</p>

<p>If used with <b>--pipe</b> this is slow.</p>

</dd>
<dt id="noswap"><b>--noswap</b></dt>
<dd>

<p>Do not start new jobs on a given computer if there is both swap-in and swap-out activity.</p>

<p>The swap activity is only sampled every 10 seconds as the sampling takes 1 second to do.</p>

<p>Swap activity is computed as (swap-in)*(swap-out) which in practice is a good value: swapping out is not a problem, swapping in is not a problem, but both swapping in and out usually indicates a problem.</p>

</dd>
<dt id="record-env"><b>--record-env</b></dt>
<dd>

<p>Record current environment variables in ~/.parallel/ignored_vars. This is useful before using <b>--env _</b>.</p>

<p>See also <b>--env</b>.</p>

</dd>
<dt id="recstart-startstring"><b>--recstart</b> <i>startstring</i></dt>
<dd>

</dd>
<dt id="recend-endstring"><b>--recend</b> <i>endstring</i></dt>
<dd>

<p>If <b>--recstart</b> is given <i>startstring</i> will be used to split at record start.</p>

<p>If <b>--recend</b> is given <i>endstring</i> will be used to split at record end.</p>

<p>If both <b>--recstart</b> and <b>--recend</b> are given the combined string <i>endstring</i><i>startstring</i> will have to match to find a split position. This is useful if either <i>startstring</i> or <i>endstring</i> match in the middle of a record.</p>

<p>If neither <b>--recstart</b> nor <b>--recend</b> are given then <b>--recend</b> defaults to &#39;\n&#39;. To have no record separator use <b>--recend &quot;&quot;</b>.</p>

<p><b>--recstart</b> and <b>--recend</b> are used with <b>--pipe</b>.</p>

<p>Use <b>--regexp</b> to interpret <b>--recstart</b> and <b>--recend</b> as regular expressions. This is slow, however.</p>

</dd>
<dt id="regexp"><b>--regexp</b></dt>
<dd>

<p>Use <b>--regexp</b> to interpret <b>--recstart</b> and <b>--recend</b> as regular expressions. This is slow, however.</p>

</dd>
<dt id="remove-rec-sep"><b>--remove-rec-sep</b></dt>
<dd>

</dd>
<dt id="removerecsep"><b>--removerecsep</b></dt>
<dd>

</dd>
<dt id="rrs"><b>--rrs</b></dt>
<dd>

<p>Remove the text matched by <b>--recstart</b> and <b>--recend</b> before piping it to the command.</p>

<p>Only used with <b>--pipe</b>.</p>

</dd>
<dt id="results-prefix"><b>--results</b> <i>prefix</i></dt>
<dd>

</dd>
<dt id="res-prefix"><b>--res</b> <i>prefix</i></dt>
<dd>

<p>Save the output into files. The files will be stored in a directory tree rooted at <i>prefix</i>. Within this directory tree, each command will result in two files: <i>prefix</i>/&lt;ARGS&gt;/stdout and <i>prefix</i>/&lt;ARGS&gt;/stderr, where &lt;ARGS&gt; is a sequence of directories representing the header of the input source (if using <b>--header :</b>) or the number of the input source and corresponding values.</p>

<p>E.g:</p>

<pre><code>  parallel --header : --results foo echo {a} {b} ::: a I II ::: b III IIII</code></pre>

<p>will generate the files:</p>

<pre><code>  foo/a/I/b/III/stderr
  foo/a/I/b/III/stdout
  foo/a/I/b/IIII/stderr
  foo/a/I/b/IIII/stdout
  foo/a/II/b/III/stderr
  foo/a/II/b/III/stdout
  foo/a/II/b/IIII/stderr
  foo/a/II/b/IIII/stdout</code></pre>

<p>and</p>

<pre><code>  parallel --results foo echo {1} {2} ::: I II ::: III IIII</code></pre>

<p>will generate the files:</p>

<pre><code>  foo/1/I/2/III/stderr
  foo/1/I/2/III/stdout
  foo/1/I/2/IIII/stderr
  foo/1/I/2/IIII/stdout
  foo/1/II/2/III/stderr
  foo/1/II/2/III/stdout
  foo/1/II/2/IIII/stderr
  foo/1/II/2/IIII/stdout</code></pre>

<p>See also <b>--files</b>, <b>--header</b>, <b>--joblog</b>.</p>

</dd>
<dt id="resume"><b>--resume</b></dt>
<dd>

<p>Resumes from the last unfinished job. By reading <b>--joblog</b> or the <b>--results</b> dir GNU <b>parallel</b> will figure out the last unfinished job and continue from there. As GNU <b>parallel</b> only looks at the sequence numbers in <b>--joblog</b> then the input, the command, and <b>--joblog</b> all have to remain unchanged; otherwise GNU <b>parallel</b> may run wrong commands.</p>

<p>See also <b>--joblog</b>, <b>--results</b>, <b>--resume-failed</b>.</p>

</dd>
<dt id="resume-failed"><b>--resume-failed</b></dt>
<dd>

<p>Retry all failed and resume from the last unfinished job. By reading <b>--joblog</b> GNU <b>parallel</b> will figure out the failed jobs and run those again. After that it will resume last unfinished job and continue from there. As GNU <b>parallel</b> only looks at the sequence numbers in <b>--joblog</b> then the input, the command, and <b>--joblog</b> all have to remain unchanged; otherwise GNU <b>parallel</b> may run wrong commands.</p>

<p>See also <b>--joblog</b>, <b>--resume</b>.</p>

</dd>
<dt id="retries-n"><b>--retries</b> <i>n</i></dt>
<dd>

<p>If a job fails, retry it on another computer on which it has not failed. Do this <i>n</i> times. If there are fewer than <i>n</i> computers in <b>--sshlogin</b> GNU <b>parallel</b> will re-use all the computers. This is useful if some jobs fail for no apparent reason (such as network failure).</p>

</dd>
<dt id="return-filename"><b>--return</b> <i>filename</i></dt>
<dd>

<p>Transfer files from remote computers. <b>--return</b> is used with <b>--sshlogin</b> when the arguments are files on the remote computers. When processing is done the file <i>filename</i> will be transferred from the remote computer using <b>rsync</b> and will be put relative to the default login dir. E.g.</p>

<pre><code>  echo foo/bar.txt | parallel \
    --sshlogin server.example.com --return {.}.out touch {.}.out</code></pre>

<p>This will transfer the file <i>$HOME/foo/bar.out</i> from the computer <i>server.example.com</i> to the file <i>foo/bar.out</i> after running <b>touch foo/bar.out</b> on <i>server.example.com</i>.</p>

<pre><code>  echo /tmp/foo/bar.txt | parallel \
    --sshlogin server.example.com --return {.}.out touch {.}.out</code></pre>

<p>This will transfer the file <i>/tmp/foo/bar.out</i> from the computer <i>server.example.com</i> to the file <i>/tmp/foo/bar.out</i> after running <b>touch /tmp/foo/bar.out</b> on <i>server.example.com</i>.</p>

<p>Multiple files can be transferred by repeating the options multiple times:</p>

<pre><code>  echo /tmp/foo/bar.txt | \
    parallel --sshlogin server.example.com \
    --return {.}.out --return {.}.out2 touch {.}.out {.}.out2</code></pre>

<p><b>--return</b> is often used with <b>--transfer</b> and <b>--cleanup</b>.</p>

<p><b>--return</b> is ignored when used with <b>--sshlogin :</b> or when not used with <b>--sshlogin</b>.</p>

</dd>
<dt id="round-robin"><b>--round-robin</b></dt>
<dd>

</dd>
<dt id="round"><b>--round</b></dt>
<dd>

<p>Normally <b>--pipe</b> will give a single block to each instance of the command. With <b>--round-robin</b> all blocks will at random be written to commands already running. This is useful if the command takes a long time to initialize.</p>

<p><b>--keep-order</b> will not work with <b>--round-robin</b> as it is impossible to track which input block corresponds to which output.</p>

</dd>
<dt id="rpl-tag-perl-expression"><b>--rpl</b> &#39;<i>tag</i> <i>perl expression</i>&#39;</dt>
<dd>

<p>Use <i>tag</i> as a replacement string for <i>perl expression</i>. This makes it possible to define your own replacement strings. GNU <b>parallel</b>&#39;s 7 replacement strings are implemented as:</p>

<pre><code>  --rpl &#39;{} &#39;
  --rpl &#39;{#} 1 $_=$job-&gt;seq()&#39;
  --rpl &#39;{%} 1 $_=$job-&gt;slot()&#39;
  --rpl &#39;{/} s:.*/::&#39;
  --rpl &#39;{//} $Global::use{&quot;File::Basename&quot;} ||= eval &quot;use File::Basename; 1;&quot;; $_ = dirname($_);&#39;
  --rpl &#39;{/.} s:.*/::; s:\.[^/.]+$::;&#39;
  --rpl &#39;{.} s:\.[^/.]+$::&#39;</code></pre>

<p>If the user defined replacement string starts with &#39;{&#39; it can also be used as a positional replacement string (like <b>{2.}</b>).</p>

<p>It is recommended to only change $_ but you have full access to all of GNU <b>parallel</b>&#39;s internal functions and data structures.</p>

<p>Here are a few examples:</p>

<pre><code>  Remove 2 extensions (e.g. .tar.gz)
  --rpl &#39;{..} s:\.[^/.]+$::;s:\.[^/.]+$::;&#39;
  Keep only the extension
  --rpl &#39;{ext} s:.*\.::&#39;
  Is the job sequence even or odd?
  --rpl &#39;{odd} $_=$job-&gt;seq()%2?&quot;odd&quot;:&quot;even&quot;&#39;</code></pre>

<p>See also: <b>{= perl expression =}</b> <b>--parens</b></p>

</dd>
<dt id="max-chars-max-chars"><b>--max-chars</b>=<i>max-chars</i></dt>
<dd>

</dd>
<dt id="s-max-chars"><b>-s</b> <i>max-chars</i></dt>
<dd>

<p>Use at most <i>max-chars</i> characters per command line, including the command and initial-arguments and the terminating nulls at the ends of the argument strings. The largest allowed value is system-dependent, and is calculated as the argument length limit for exec, less the size of your environment. The default value is the maximum.</p>

<p>Implies <b>-X</b> unless <b>-m</b> is set.</p>

</dd>
<dt id="show-limits"><b>--show-limits</b></dt>
<dd>

<p>Display the limits on the command-line length which are imposed by the operating system and the <b>-s</b> option. Pipe the input from /dev/null (and perhaps specify --no-run-if-empty) if you don&#39;t want GNU <b>parallel</b> to do anything.</p>

</dd>
<dt id="semaphore"><b>--semaphore</b></dt>
<dd>

<p>Work as a counting semaphore. <b>--semaphore</b> will cause GNU <b>parallel</b> to start <i>command</i> in the background. When the number of simultaneous jobs is reached, GNU <b>parallel</b> will wait for one of these to complete before starting another command.</p>

<p><b>--semaphore</b> implies <b>--bg</b> unless <b>--fg</b> is specified.</p>

<p><b>--semaphore</b> implies <b>--semaphorename `tty`</b> unless <b>--semaphorename</b> is specified.</p>

<p>Used with <b>--fg</b>, <b>--wait</b>, and <b>--semaphorename</b>.</p>

<p>The command <b>sem</b> is an alias for <b>parallel --semaphore</b>.</p>

<p>See also <b>man sem</b>.</p>

</dd>
<dt id="semaphorename-name"><b>--semaphorename</b> <i>name</i></dt>
<dd>

</dd>
<dt id="id-name"><b>--id</b> <i>name</i></dt>
<dd>

<p>Use <b>name</b> as the name of the semaphore. Default is the name of the controlling tty (output from <b>tty</b>).</p>

<p>The default normally works as expected when used interactively, but when used in a script <i>name</i> should be set. <i>$$</i> or <i>my_task_name</i> are often a good value.</p>

<p>The semaphore is stored in ~/.parallel/semaphores/</p>

<p>Implies <b>--semaphore</b>.</p>

<p>See also <b>man sem</b>.</p>

</dd>
<dt id="semaphoretimeout-secs"><b>--semaphoretimeout</b> <i>secs</i></dt>
<dd>

<p>If the semaphore is not released within secs seconds, take it anyway.</p>

<p>Implies <b>--semaphore</b>.</p>

<p>See also <b>man sem</b>.</p>

</dd>
<dt id="seqreplace-replace-str"><b>--seqreplace</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of <b>{#}</b> for job sequence number.</p>

</dd>
<dt id="shebang"><b>--shebang</b></dt>
<dd>

</dd>
<dt id="hashbang"><b>--hashbang</b></dt>
<dd>

<p>GNU <b>parallel</b> can be called as a shebang (#!) command as the first line of a script. The content of the file will be treated as inputsource.</p>

<p>Like this:</p>

<pre><code>  #!/usr/bin/parallel --shebang -r traceroute

  foss.org.my
  debian.org
  freenetproject.org</code></pre>

<p><b>--shebang</b> must be set as the first option.</p>

<p>On FreeBSD <b>env</b> is needed:</p>

<pre><code>  #!/usr/bin/env -S parallel --shebang -r traceroute

  foss.org.my
  debian.org
  freenetproject.org</code></pre>

</dd>
<dt id="shebang-wrap"><b>--shebang-wrap</b></dt>
<dd>

<p>GNU <b>parallel</b> can parallelize scripts by wrapping the shebang line. If the program can be run like this:</p>

<pre><code>  cat arguments | parallel the_program</code></pre>

<p>then the script can be changed to:</p>

<pre><code>  #!/usr/bin/parallel --shebang-wrap /the/original/parser --with-options</code></pre>

<p>E.g.</p>

<pre><code>  #!/usr/bin/parallel --shebang-wrap /usr/bin/python</code></pre>

<p>If the program can be run like this:</p>

<pre><code>  cat data | parallel --pipe the_program</code></pre>

<p>then the script can be changed to:</p>

<pre><code>  #!/usr/bin/parallel --shebang-wrap --pipe /the/original/parser --with-options</code></pre>

<p>E.g.</p>

<pre><code>  #!/usr/bin/parallel --shebang-wrap --pipe /usr/bin/perl -w</code></pre>

<p><b>--shebang-wrap</b> must be set as the first option.</p>

</dd>
<dt id="shellquote"><b>--shellquote</b></dt>
<dd>

<p>Does not run the command but quotes it. Useful for making quoted composed commands for GNU <b>parallel</b>.</p>

</dd>
<dt id="skip-first-line"><b>--skip-first-line</b></dt>
<dd>

<p>Do not use the first line of input (used by GNU <b>parallel</b> itself when called with <b>--shebang</b>).</p>

</dd>
<dt id="sshdelay-secs"><b>--sshdelay</b> <i>secs</i></dt>
<dd>

<p>Delay starting next ssh by <i>secs</i> seconds. GNU <b>parallel</b> will pause <i>secs</i> seconds after starting each ssh. <i>secs</i> can be less than 1 seconds.</p>

</dd>
<dt id="S-ncpu-sshlogin-ncpu-sshlogin"><b>-S</b> <i>[ncpu/]sshlogin[,[ncpu/]sshlogin[,...]]</i></dt>
<dd>

</dd>
<dt id="sshlogin-ncpu-sshlogin-ncpu-sshlogin"><b>--sshlogin</b> <i>[ncpu/]sshlogin[,[ncpu/]sshlogin[,...]]</i></dt>
<dd>

<p>Distribute jobs to remote computers. The jobs will be run on a list of remote computers. GNU <b>parallel</b> will determine the number of CPU cores on the remote computers and run the number of jobs as specified by <b>-j</b>. If the number <i>ncpu</i> is given GNU <b>parallel</b> will use this number for number of CPU cores on the host. Normally <i>ncpu</i> will not be needed.</p>

<p>An <i>sshlogin</i> is of the form:</p>

<pre><code>  [sshcommand [options]] [username@]hostname</code></pre>

<p>The sshlogin must not require a password.</p>

<p>The sshlogin &#39;:&#39; is special, it means &#39;no ssh&#39; and will therefore run on the local computer.</p>

<p>The sshlogin &#39;..&#39; is special, it read sshlogins from ~/.parallel/sshloginfile</p>

<p>The sshlogin &#39;-&#39; is special, too, it read sshlogins from stdin (standard input).</p>

<p>To specify more sshlogins separate the sshlogins by comma or repeat the options multiple times.</p>

<p>For examples: see <b>--sshloginfile</b>.</p>

<p>The remote host must have GNU <b>parallel</b> installed.</p>

<p><b>--sshlogin</b> is known to cause problems with <b>-m</b> and <b>-X</b>.</p>

<p><b>--sshlogin</b> is often used with <b>--transfer</b>, <b>--return</b>, <b>--cleanup</b>, and <b>--trc</b>.</p>

</dd>
<dt id="sshloginfile-filename-beta-testing"><b>--sshloginfile</b> <i>filename</i> (beta testing)</dt>
<dd>

</dd>
<dt id="slf-filename-beta-testing"><b>--slf</b> <i>filename</i> (beta testing)</dt>
<dd>

<p>File with sshlogins. The file consists of sshlogins on separate lines. Empty lines and lines starting with &#39;#&#39; are ignored. Example:</p>

<pre><code>  server.example.com
  username@server2.example.com
  8/my-8-core-server.example.com
  2/my_other_username@my-dualcore.example.net
  # This server has SSH running on port 2222
  ssh -p 2222 server.example.net
  4/ssh -p 2222 quadserver.example.net
  # Use a different ssh program
  myssh -p 2222 -l myusername hexacpu.example.net
  # Use a different ssh program with default number of cores
  //usr/local/bin/myssh -p 2222 -l myusername hexacpu.example.net
  # Use a different ssh program with 6 cores
  6//usr/local/bin/myssh -p 2222 -l myusername hexacpu.example.net
  # Assume 16 cores on the local computer
  16/:</code></pre>

<p>When using a different ssh program the last argument must be the hostname.</p>

<p>Multiple <b>--sshloginfile</b> are allowed.</p>

<p>GNU <b>parallel</b> will first look for the file in current dir; if that fails it look for the file in ~/.parallel.</p>

<p>The sshloginfile &#39;..&#39; is special, it read sshlogins from ~/.parallel/sshloginfile</p>

<p>The sshloginfile &#39;.&#39; is special, it read sshlogins from /etc/parallel/sshloginfile</p>

<p>The sshloginfile &#39;-&#39; is special, too, it read sshlogins from stdin (standard input).</p>

<p>If the sshloginfile is changed it will be re-read when a job finishes though at most once per second. This makes it possible to add and remove hosts while running.</p>

<p>This can be used to have a daemon that updates the sshloginfile to only contain servers that are up:</p>

<pre><code>    cp original.slf tmp2.slf
    while [ 1 ] ; do
      nice parallel --nonall -j0 -k --slf original.slf --tag echo | perl &#39;s/\t$//&#39; &gt; tmp.slf
      if diff tmp.slf tmp2.slf; then
        mv tmp.slf tmp2.slf
      fi
      sleep 10
    done &amp;
    parallel --slf tmp2.slf ...</code></pre>

</dd>
<dt id="slotreplace-replace-str"><b>--slotreplace</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of <b>{%}</b> for job slot number.</p>

</dd>
<dt id="silent"><b>--silent</b></dt>
<dd>

<p>Silent. The job to be run will not be printed. This is the default. Can be reversed with <b>-v</b>.</p>

</dd>
<dt id="tty"><b>--tty</b></dt>
<dd>

<p>Open terminal tty. If GNU <b>parallel</b> is used for starting an interactive program then this option may be needed. It will start only one job at a time (i.e. <b>-j1</b>), not buffer the output (i.e. <b>-u</b>), and it will open a tty for the job. When the job is done, the next job will get the tty.</p>

</dd>
<dt id="tag"><b>--tag</b></dt>
<dd>

<p>Tag lines with arguments. Each output line will be prepended with the arguments and TAB (\t). When combined with <b>--onall</b> or <b>--nonall</b> the lines will be prepended with the sshlogin instead.</p>

<p><b>--tag</b> is ignored when using <b>-u</b>.</p>

</dd>
<dt id="tagstring-str"><b>--tagstring</b> <i>str</i></dt>
<dd>

<p>Tag lines with a string. Each output line will be prepended with <i>str</i> and TAB (\t). <i>str</i> can contain replacement strings such as {}.</p>

<p><b>--tagstring</b> is ignored when using <b>-u</b>, <b>--onall</b>, and <b>--nonall</b>.</p>

</dd>
<dt id="tmpdir-dirname"><b>--tmpdir</b> <i>dirname</i></dt>
<dd>

<p>Directory for temporary files. GNU <b>parallel</b> normally buffers output into temporary files in /tmp. By setting <b>--tmpdir</b> you can use a different dir for the files. Setting <b>--tmpdir</b> is equivalent to setting $TMPDIR.</p>

</dd>
<dt id="tmux"><b>--tmux</b></dt>
<dd>

<p>Use <b>tmux</b> for output. Start a <b>tmux</b> session and run each job in a window in that session. No other output will be produced.</p>

</dd>
<dt id="timeout-val"><b>--timeout</b> <i>val</i></dt>
<dd>

<p>Time out for command. If the command runs for longer than <i>val</i> seconds it will get killed with SIGTERM, followed by SIGTERM 200 ms later, followed by SIGKILL 200 ms later.</p>

<p>If <i>val</i> is followed by a % then the timeout will dynamically be computed as a percentage of the median average runtime. Only values &gt; 100% will make sense.</p>

</dd>
<dt id="verbose"><b>--verbose</b></dt>
<dd>

</dd>
<dt id="t"><b>-t</b></dt>
<dd>

<p>Print the job to be run on stderr (standard error).</p>

<p>See also <b>-v</b>, <b>-p</b>.</p>

</dd>
<dt id="transfer"><b>--transfer</b></dt>
<dd>

<p>Transfer files to remote computers. <b>--transfer</b> is used with <b>--sshlogin</b> when the arguments are files and should be transferred to the remote computers. The files will be transferred using <b>rsync</b> and will be put relative to the default work dir. If the path contains /./ the remaining path will be relative to the work dir. E.g.</p>

<pre><code>  echo foo/bar.txt | parallel \
    --sshlogin server.example.com --transfer wc</code></pre>

<p>This will transfer the file <i>foo/bar.txt</i> to the computer <i>server.example.com</i> to the file <i>$HOME/foo/bar.txt</i> before running <b>wc foo/bar.txt</b> on <i>server.example.com</i>.</p>

<pre><code>  echo /tmp/foo/bar.txt | parallel \
    --sshlogin server.example.com --transfer wc</code></pre>

<p>This will transfer the file <i>foo/bar.txt</i> to the computer <i>server.example.com</i> to the file <i>/tmp/foo/bar.txt</i> before running <b>wc /tmp/foo/bar.txt</b> on <i>server.example.com</i>.</p>

<p><b>--transfer</b> is often used with <b>--return</b> and <b>--cleanup</b>.</p>

<p><b>--transfer</b> is ignored when used with <b>--sshlogin :</b> or when not used with <b>--sshlogin</b>.</p>

</dd>
<dt id="trc-filename"><b>--trc</b> <i>filename</i></dt>
<dd>

<p>Transfer, Return, Cleanup. Short hand for:</p>

<p><b>--transfer</b> <b>--return</b> <i>filename</i> <b>--cleanup</b></p>

</dd>
<dt id="trim-n-l-r-lr-rl"><b>--trim</b> &lt;n|l|r|lr|rl&gt;</dt>
<dd>

<p>Trim white space in input.</p>

<dl>

<dt id="n5">n</dt>
<dd>

<p>No trim. Input is not modified. This is the default.</p>

</dd>
<dt id="l">l</dt>
<dd>

<p>Left trim. Remove white space from start of input. E.g. &quot; a bc &quot; -&gt; &quot;a bc &quot;.</p>

</dd>
<dt id="r1">r</dt>
<dd>

<p>Right trim. Remove white space from end of input. E.g. &quot; a bc &quot; -&gt; &quot; a bc&quot;.</p>

</dd>
<dt id="lr">lr</dt>
<dd>

</dd>
<dt id="rl">rl</dt>
<dd>

<p>Both trim. Remove white space from both start and end of input. E.g. &quot; a bc &quot; -&gt; &quot;a bc&quot;. This is the default if <b>--colsep</b> is used.</p>

</dd>
</dl>

</dd>
<dt id="ungroup"><b>--ungroup</b></dt>
<dd>

</dd>
<dt id="u"><b>-u</b></dt>
<dd>

<p>Ungroup output. Output is printed as soon as possible and by passes GNU <b>parallel</b> internal processing. This may cause output from different commands to be mixed thus should only be used if you do not care about the output. Compare these:</p>

<p><b>parallel -j0 &#39;sleep {};echo -n start{};sleep {};echo {}end&#39; ::: 1 2 3 4</b></p>

<p><b>parallel -u -j0 &#39;sleep {};echo -n start{};sleep {};echo {}end&#39; ::: 1 2 3 4</b></p>

<p>It also disables <b>--tag</b>. GNU <b>parallel</b> outputs faster with <b>-u</b>. Compare the speed of these:</p>

<pre><code>  parallel seq ::: 300000000 &gt;/dev/null
  parallel -u seq ::: 300000000 &gt;/dev/null
  parallel --line-buffer seq ::: 300000000 &gt;/dev/null</code></pre>

<p>Can be reversed with <b>--group</b>.</p>

<p>See also: <b>--line-buffer</b> <b>--group</b></p>

</dd>
<dt id="extensionreplace-replace-str"><b>--extensionreplace</b> <i>replace-str</i></dt>
<dd>

</dd>
<dt id="er-replace-str"><b>--er</b> <i>replace-str</i></dt>
<dd>

<p>Use the replacement string <i>replace-str</i> instead of {.} for input line without extension.</p>

</dd>
<dt id="use-cpus-instead-of-cores"><b>--use-cpus-instead-of-cores</b></dt>
<dd>

<p>Count the number of physical CPUs instead of CPU cores. When computing how many jobs to run simultaneously relative to the number of CPU cores you can ask GNU <b>parallel</b> to instead look at the number of physical CPUs. This will make sense for computers that have hyperthreading as two jobs running on one CPU with hyperthreading will run slower than two jobs running on two physical CPUs. Some multi-core CPUs can run faster if only one thread is running per physical CPU. Most users will not need this option.</p>

</dd>
<dt id="v"><b>-v</b></dt>
<dd>

<p>Verbose. Print the job to be run on stdout (standard output). Can be reversed with <b>--silent</b>. See also <b>-t</b>.</p>

<p>Use <b>-v</b> <b>-v</b> to print the wrapping ssh command when running remotely.</p>

</dd>
<dt id="version"><b>--version</b></dt>
<dd>

</dd>
<dt id="V"><b>-V</b></dt>
<dd>

<p>Print the version GNU <b>parallel</b> and exit.</p>

</dd>
<dt id="workdir-mydir"><b>--workdir</b> <i>mydir</i></dt>
<dd>

</dd>
<dt id="wd-mydir"><b>--wd</b> <i>mydir</i></dt>
<dd>

<p>Files transferred using <b>--transfer</b> and <b>--return</b> will be relative to <i>mydir</i> on remote computers, and the command will be executed in the dir <i>mydir</i>.</p>

<p>The special <i>mydir</i> value <b>...</b> will create working dirs under <b>~/.parallel/tmp/</b> on the remote computers. If <b>--cleanup</b> is given these dirs will be removed.</p>

<p>The special <i>mydir</i> value <b>.</b> uses the current working dir. If the current working dir is beneath your home dir, the value <b>.</b> is treated as the relative path to your home dir. This means that if your home dir is different on remote computers (e.g. if your login is different) the relative path will still be relative to your home dir.</p>

<p>To see the difference try:</p>

<p><b>parallel -S server pwd ::: &quot;&quot;</b></p>

<p><b>parallel --wd . -S server pwd ::: &quot;&quot;</b></p>

<p><b>parallel --wd ... -S server pwd ::: &quot;&quot;</b></p>

</dd>
<dt id="wait"><b>--wait</b></dt>
<dd>

<p>Wait for all commands to complete.</p>

<p>Implies <b>--semaphore</b>.</p>

<p>See also <b>man sem</b>.</p>

</dd>
<dt id="X"><b>-X</b></dt>
<dd>

<p>Multiple arguments with context replace. Insert as many arguments as the command line length permits. If multiple jobs are being run in parallel: distribute the arguments evenly among the jobs. Use <b>-j1</b> to avoid this.</p>

<p>If <b>{}</b> is not used the arguments will be appended to the line. If <b>{}</b> is used as part of a word (like <i>pic{}.jpg</i>) then the whole word will be repeated. If <b>{}</b> is used multiple times each <b>{}</b> will be replaced with the arguments.</p>

<p>Normally <b>-X</b> will do the right thing, whereas <b>-m</b> can give unexpected results if <b>{}</b> is used as part of a word.</p>

<p>Support for <b>-X</b> with <b>--sshlogin</b> is limited and may fail.</p>

<p>See also <b>-m</b>.</p>

</dd>
<dt id="exit"><b>--exit</b></dt>
<dd>

</dd>
<dt id="x"><b>-x</b></dt>
<dd>

<p>Exit if the size (see the <b>-s</b> option) is exceeded.</p>

</dd>
<dt id="xapply"><b>--xapply</b></dt>
<dd>

<p>Read multiple input sources like <b>xapply</b>. If multiple input sources are given, one argument will be read from each of the input sources. The arguments can be accessed in the command as <b>{1}</b> .. <b>{</b><i>n</i><b>}</b>, so <b>{1}</b> will be a line from the first input source, and <b>{6}</b> will refer to the line with the same line number from the 6th input source.</p>

<p>Compare these two:</p>

<pre><code>  parallel echo {1} {2} ::: 1 2 3 ::: a b c
  parallel --xapply echo {1} {2} ::: 1 2 3 ::: a b c</code></pre>

<p>Arguments will be recycled if one input source has more arguments than the others:</p>

<pre><code>  parallel --xapply echo {1} {2} {3} ::: 1 2 ::: I II III ::: a b c d e f g</code></pre>

<p>See also <b>--header</b>.</p>

</dd>
</dl>

<h1 id="EXAMPLE:-Working-as-xargs--n1.-Argument-appending">EXAMPLE: Working as xargs -n1. Argument appending</h1>

<p>GNU <b>parallel</b> can work similar to <b>xargs -n1</b>.</p>

<p>To compress all html files using <b>gzip</b> run:</p>

<p><b>find . -name &#39;*.html&#39; | parallel gzip --best</b></p>

<p>If the file names may contain a newline use <b>-0</b>. Substitute FOO BAR with FUBAR in all files in this dir and subdirs:</p>

<p><b>find . -type f -print0 | parallel -q0 perl -i -pe &#39;s/FOO BAR/FUBAR/g&#39;</b></p>

<p>Note <b>-q</b> is needed because of the space in &#39;FOO BAR&#39;.</p>

<h1 id="EXAMPLE:-Reading-arguments-from-command-line">EXAMPLE: Reading arguments from command line</h1>

<p>GNU <b>parallel</b> can take the arguments from command line instead of stdin (standard input). To compress all html files in the current dir using <b>gzip</b> run:</p>

<p><b>parallel gzip --best ::: *.html</b></p>

<p>To convert *.wav to *.mp3 using LAME running one process per CPU core run:</p>

<p><b>parallel lame {} -o {.}.mp3 ::: *.wav</b></p>

<h1 id="EXAMPLE:-Inserting-multiple-arguments">EXAMPLE: Inserting multiple arguments</h1>

<p>When moving a lot of files like this: <b>mv *.log destdir</b> you will sometimes get the error:</p>

<p><b>bash: /bin/mv: Argument list too long</b></p>

<p>because there are too many files. You can instead do:</p>

<p><b>ls | grep -E &#39;\.log$&#39; | parallel mv {} destdir</b></p>

<p>This will run <b>mv</b> for each file. It can be done faster if <b>mv</b> gets as many arguments that will fit on the line:</p>

<p><b>ls | grep -E &#39;\.log$&#39; | parallel -m mv {} destdir</b></p>

<h1 id="EXAMPLE:-Context-replace">EXAMPLE: Context replace</h1>

<p>To remove the files <i>pict0000.jpg</i> .. <i>pict9999.jpg</i> you could do:</p>

<p><b>seq -w 0 9999 | parallel rm pict{}.jpg</b></p>

<p>You could also do:</p>

<p><b>seq -w 0 9999 | perl -pe &#39;s/(.*)/pict$1.jpg/&#39; | parallel -m rm</b></p>

<p>The first will run <b>rm</b> 10000 times, while the last will only run <b>rm</b> as many times needed to keep the command line length short enough to avoid <b>Argument list too long</b> (it typically runs 1-2 times).</p>

<p>You could also run:</p>

<p><b>seq -w 0 9999 | parallel -X rm pict{}.jpg</b></p>

<p>This will also only run <b>rm</b> as many times needed to keep the command line length short enough.</p>

<h1 id="EXAMPLE:-Compute-intensive-jobs-and-substitution">EXAMPLE: Compute intensive jobs and substitution</h1>

<p>If ImageMagick is installed this will generate a thumbnail of a jpg file:</p>

<p><b>convert -geometry 120 foo.jpg thumb_foo.jpg</b></p>

<p>This will run with number-of-cpu-cores jobs in parallel for all jpg files in a directory:</p>

<p><b>ls *.jpg | parallel convert -geometry 120 {} thumb_{}</b></p>

<p>To do it recursively use <b>find</b>:</p>

<p><b>find . -name &#39;*.jpg&#39; | parallel convert -geometry 120 {} {}_thumb.jpg</b></p>

<p>Notice how the argument has to start with <b>{}</b> as <b>{}</b> will include path (e.g. running <b>convert -geometry 120 ./foo/bar.jpg thumb_./foo/bar.jpg</b> would clearly be wrong). The command will generate files like ./foo/bar.jpg_thumb.jpg.</p>

<p>Use <b>{.}</b> to avoid the extra .jpg in the file name. This command will make files like ./foo/bar_thumb.jpg:</p>

<p><b>find . -name &#39;*.jpg&#39; | parallel convert -geometry 120 {} {.}_thumb.jpg</b></p>

<h1 id="EXAMPLE:-Substitution-and-redirection">EXAMPLE: Substitution and redirection</h1>

<p>This will generate an uncompressed version of .gz-files next to the .gz-file:</p>

<p><b>parallel zcat {} &quot;</b>&gt;<b>&quot;{.} ::: *.gz</b></p>

<p>Quoting of &gt; is necessary to postpone the redirection. Another solution is to quote the whole command:</p>

<p><b>parallel &quot;zcat {} </b>&gt;<b>{.}&quot; ::: *.gz</b></p>

<p>Other special shell characters (such as * ; $ &gt; &lt; | &gt;&gt; &lt;&lt;) also need to be put in quotes, as they may otherwise be interpreted by the shell and not given to GNU <b>parallel</b>.</p>

<h1 id="EXAMPLE:-Composed-commands">EXAMPLE: Composed commands</h1>

<p>A job can consist of several commands. This will print the number of files in each directory:</p>

<p><b>ls | parallel &#39;echo -n {}&quot; &quot;; ls {}|wc -l&#39;</b></p>

<p>To put the output in a file called &lt;name&gt;.dir:</p>

<p><b>ls | parallel &#39;(echo -n {}&quot; &quot;; ls {}|wc -l) </b>&gt; <b>{}.dir&#39;</b></p>

<p>Even small shell scripts can be run by GNU <b>parallel</b>:</p>

<p><b>find . | parallel &#39;a={}; name=${a##*/}; upper=$(echo &quot;$name&quot; | tr &quot;[:lower:]&quot; &quot;[:upper:]&quot;); echo &quot;$name - $upper&quot;&#39;</b></p>

<p><b>ls | parallel &#39;mv {} &quot;$(echo {} | tr &quot;[:upper:]&quot; &quot;[:lower:]&quot;)&quot;&#39;</b></p>

<p>Given a list of URLs, list all URLs that fail to download. Print the line number and the URL.</p>

<p><b>cat urlfile | parallel &quot;wget {} 2</b>&gt;<b>/dev/null || grep -n {} urlfile&quot;</b></p>

<p>Create a mirror directory with the same filenames except all files and symlinks are empty files.</p>

<p><b>cp -rs /the/source/dir mirror_dir; find mirror_dir -type l | parallel -m rm {} &#39;&amp;&amp;&#39; touch {}</b></p>

<p>Find the files in a list that do not exist</p>

<p><b>cat file_list | parallel &#39;if [ ! -e {} ] ; then echo {}; fi&#39;</b></p>

<h1 id="EXAMPLE:-Calling-Bash-functions">EXAMPLE: Calling Bash functions</h1>

<p>If the composed command is longer than a line, it becomes hard to read. In Bash you can use functions. Just remember to <b>export -f</b> the function.</p>

<pre><code>  doit() {
    echo Doing it for $1
    sleep 2
    echo Done with $1
  }
  export -f doit
  parallel doit ::: 1 2 3

  doubleit() {
    echo Doing it for $1 $2
    sleep 2
    echo Done with $1 $2
  }
  export -f doubleit
  parallel doubleit ::: 1 2 3 ::: a b</code></pre>

<p>To do this on remote servers you need to transfer the function using <b>--env</b>:</p>

<pre><code>  parallel --env doit -S server doit ::: 1 2 3
  parallel --env doubleit -S server doubleit ::: 1 2 3 ::: a b</code></pre>

<p>If your environment (aliases, variables, and functions) is small you can copy the full environment without having to <b>export -f</b> anything. Just run this first:</p>

<pre><code>  env_parallel() {
      export parallel_bash_environment=&#39;() {
        &#39;&quot;$(echo &quot;shopt -s expand_aliases 2&gt;/dev/null&quot;; alias;typeset -p | grep -vFf &lt;(readonly; echo GROUPS; echo FUNCNAME; echo DIRSTACK; echo _; echo PIPESTATUS; echo USERNAME) | grep -v BASH_;typeset -f)&quot;&#39;
      }&#39;
      # Run as: env_parallel parallel_bash_environment &quot;2&gt;/dev/null;&quot; ...
      `which parallel` &quot;$@&quot;
      unset parallel_bash_environment
  }</code></pre>

<p>And then call as:</p>

<pre><code>  env_parallel doit ::: 1 2 3
  env_parallel doubleit ::: 1 2 3 ::: a b
  env_parallel -S server doit ::: 1 2 3
  env_parallel -S server doubleit ::: 1 2 3 ::: a b</code></pre>

<h1 id="EXAMPLE:-Function-tester">EXAMPLE: Function tester</h1>

<p>To test a program with different parameters:</p>

<pre><code>  tester() {
    if (eval &quot;$@&quot;) &gt;&amp;/dev/null; then
      perl -e &#39;printf &quot;\033[30;102m[ OK ]\033[0m @ARGV\n&quot;&#39; &quot;$@&quot;
    else
      perl -e &#39;printf &quot;\033[30;101m[FAIL]\033[0m @ARGV\n&quot;&#39; &quot;$@&quot;
    fi
  }
  export -f tester
  parallel tester my_program ::: arg1 arg2
  parallel tester exit ::: 1 0 2 0</code></pre>

<p>If <b>my_program</b> fails a red FAIL will be printed followed by the failing command; otherwise a green OK will be printed followed by the command.</p>

<h1 id="EXAMPLE:-Removing-file-extension-when-processing-files">EXAMPLE: Removing file extension when processing files</h1>

<p>When processing files removing the file extension using <b>{.}</b> is often useful.</p>

<p>Create a directory for each zip-file and unzip it in that dir:</p>

<p><b>parallel &#39;mkdir {.}; cd {.}; unzip ../{}&#39; ::: *.zip</b></p>

<p>Recompress all .gz files in current directory using <b>bzip2</b> running 1 job per CPU core in parallel:</p>

<p><b>parallel &quot;zcat {} | bzip2 </b>&gt;<b>{.}.bz2 &amp;&amp; rm {}&quot; ::: *.gz</b></p>

<p>Convert all WAV files to MP3 using LAME:</p>

<p><b>find sounddir -type f -name &#39;*.wav&#39; | parallel lame {} -o {.}.mp3</b></p>

<p>Put all converted in the same directory:</p>

<p><b>find sounddir -type f -name &#39;*.wav&#39; | parallel lame {} -o mydir/{/.}.mp3</b></p>

<h1 id="EXAMPLE:-Removing-two-file-extensions-when-processing-files">EXAMPLE: Removing two file extensions when processing files</h1>

<p>If you have directory with tar.gz files and want these extracted in the corresponding dir (e.g foo.tar.gz will be extracted in the dir foo) you can do:</p>

<p><b>parallel --plus &#39;mkdir {..}; tar -C {..} -xf {}&#39; ::: *.tar.gz</b></p>

<h1 id="EXAMPLE:-Download-10-images-for-each-of-the-past-30-days">EXAMPLE: Download 10 images for each of the past 30 days</h1>

<p>Let us assume a website stores images like:</p>

<pre><code>  http://www.example.com/path/to/YYYYMMDD_##.jpg</code></pre>

<p>where YYYYMMDD is the date and ## is the number 01-10. This will download images for the past 30 days:</p>

<p><b>parallel wget http://www.example.com/path/to/&#39;$(date -d &quot;today -{1} days&quot; +%Y%m%d)_{2}.jpg&#39; ::: $(seq 30) ::: $(seq -w 10)</b></p>

<p><b>$(date -d &quot;today -{1} days&quot; +%Y%m%d)</b> will give the dates in YYYYMMDD with {1} days subtracted.</p>

<h1 id="EXAMPLE:-Copy-files-as-last-modified-date-ISO8601-with-added-random-digits">EXAMPLE: Copy files as last modified date (ISO8601) with added random digits</h1>

<p><b>find . | parallel &#39;cp {} ../destdir/{= $a = int(10000*rand); $_ = `date -r &quot;$_&quot; +%FT%T&quot;$a&quot;`; chomp; =}&#39;</b></p>

<h1 id="EXAMPLE:-Digtal-clock-with-blinking">EXAMPLE: Digtal clock with &quot;blinking&quot; :</h1>

<p>The : in a digital clock blinks. To make every other line have a &#39;:&#39; and the rest a &#39; &#39; a perl expression is used to look at the 3rd input source. If the value modudo 2 is 1: Use &quot;:&quot; otherwise use &quot; &quot;:</p>

<p><b>parallel -k echo {1}&#39;{=3 $_=$_%2?&quot;:&quot;:&quot; &quot;=}&#39;{2}{3} ::: {0..12} ::: {0..5} ::: {0..9}</b></p>

<h1 id="EXAMPLE:-Aggregating-content-of-files">EXAMPLE: Aggregating content of files</h1>

<p>This:</p>

<pre><code>  parallel --header : echo x{X}y{Y}z{Z} \&gt; x{X}y{Y}z{Z} \
  ::: X {1..5} ::: Y {01..10} ::: Z {1..5}</code></pre>

<p>will generate the files x1y01z1 .. x5y10z5. If you want to aggregate the output grouping on x and z you can do this:</p>

<pre><code>  parallel eval &#39;cat {=s/y01/y*/=} &gt; {=s/y01//=}&#39; ::: *y01*</code></pre>

<p>For all values of x and z it runs commands like:</p>

<pre><code>  cat x1y*z1 &gt; x1z1</code></pre>

<p>So you end up with x1z1 .. x1z5 each containing the content of all values of y.</p>

<h1 id="EXAMPLE:-Breadth-first-parallel-web-crawler-mirrorer">EXAMPLE: Breadth first parallel web crawler/mirrorer</h1>

<p>This script below will crawl and mirror a URL in parallel. It downloads first pages that are 1 click down, then 2 clicks down, then 3; instead of the normal depth first, where the first link link on each page is fetched first.</p>

<p>Run like this:</p>

<p><b>PARALLEL=-j100 ./parallel-crawl http://gatt.org.yeslab.org/</b></p>

<p>Remove the <b>wget</b> part if you only want a web crawler.</p>

<p>It works by fetching a page from a list of URLs and looking for links in that page that are within the same starting URL and that have not already been seen. These links are added to a new queue. When all the pages from the list is done, the new queue is moved to the list of URLs and the process is started over until no unseen links are found.</p>

<pre><code>  #!/bin/bash

  # E.g. http://gatt.org.yeslab.org/
  URL=$1
  # Stay inside the start dir
  BASEURL=$(echo $URL | perl -pe &#39;s:#.*::; s:(//.*/)[^/]*:$1:&#39;)
  URLLIST=$(mktemp urllist.XXXX)
  URLLIST2=$(mktemp urllist.XXXX)
  SEEN=$(mktemp seen.XXXX)

  # Spider to get the URLs
  echo $URL &gt;$URLLIST
  cp $URLLIST $SEEN

  while [ -s $URLLIST ] ; do
    cat $URLLIST |
      parallel lynx -listonly -image_links -dump {} \; wget -qm -l1 -Q1 {} \; echo Spidered: {} \&gt;\&amp;2 |
      perl -ne &#39;s/#.*//; s/\s+\d+.\s(\S+)$/$1/ and do { $seen{$1}++ or print }&#39; |
      grep -F $BASEURL |
      grep -v -x -F -f $SEEN | tee -a $SEEN &gt; $URLLIST2
    mv $URLLIST2 $URLLIST
  done

  rm -f $URLLIST $URLLIST2 $SEEN</code></pre>

<h1 id="EXAMPLE:-Process-files-from-a-tar-file-while-unpacking">EXAMPLE: Process files from a tar file while unpacking</h1>

<p>If the files to be processed are in a tar file then unpacking one file and processing it immediately may be faster than first unpacking all files.</p>

<p><b>tar xvf foo.tgz | perl -ne &#39;print $l;$l=$_;END{print $l}&#39; | parallel echo</b></p>

<p>The Perl one-liner is needed to avoid race condition.</p>

<h1 id="EXAMPLE:-Rewriting-a-for-loop-and-a-while-read-loop">EXAMPLE: Rewriting a for-loop and a while-read-loop</h1>

<p>for-loops like this:</p>

<pre><code>  (for x in `cat list` ; do
    do_something $x
  done) | process_output</code></pre>

<p>and while-read-loops like this:</p>

<pre><code>  cat list | (while read x ; do
    do_something $x
  done) | process_output</code></pre>

<p>can be written like this:</p>

<p><b>cat list | parallel do_something | process_output</b></p>

<p>For example: Find which host name in a list has IP address 1.2.3 4:</p>

<p><b>cat hosts.txt | parallel -P 100 host | grep 1.2.3.4</b></p>

<p>If the processing requires more steps the for-loop like this:</p>

<pre><code> (for x in `cat list` ; do
   no_extension=${x%.*};
   do_something $x scale $no_extension.jpg
   do_step2 &lt;$x $no_extension
 done) | process_output</code></pre>

<p>and while-loops like this:</p>

<pre><code> cat list | (while read x ; do
   no_extension=${x%.*};
   do_something $x scale $no_extension.jpg
   do_step2 &lt;$x $no_extension
 done) | process_output</code></pre>

<p>can be written like this:</p>

<p><b>cat list | parallel &quot;do_something {} scale {.}.jpg ; do_step2 &lt;{} {.}&quot; | process_output</b></p>

<h1 id="EXAMPLE:-Rewriting-nested-for-loops">EXAMPLE: Rewriting nested for-loops</h1>

<p>Nested for-loops like this:</p>

<pre><code>  (for x in `cat xlist` ; do
    for y in `cat ylist` ; do
      do_something $x $y
    done
  done) | process_output</code></pre>

<p>can be written like this:</p>

<p><b>parallel do_something {1} {2} :::: xlist ylist | process_output</b></p>

<p>Nested for-loops like this:</p>

<pre><code>  (for gender in M F ; do
    for size in S M L XL XXL ; do
      echo $gender $size
    done
  done) | sort</code></pre>

<p>can be written like this:</p>

<p><b>parallel echo {1} {2} ::: M F ::: S M L XL XXL | sort</b></p>

<h1 id="EXAMPLE:-Finding-the-lowest-difference-between-files">EXAMPLE: Finding the lowest difference between files</h1>

<p><b>diff</b> is good for finding differences in text files. <b>diff | wc -l</b> gives an indication of the size of the difference. To find the differences between all files in the current dir do:</p>

<p><b>parallel --tag &#39;diff {1} {2} | wc -l&#39; ::: * ::: * | sort -nk3</b></p>

<p>This way it is possible to see if some files are closer to other files.</p>

<h1 id="EXAMPLE:-for-loops-with-column-names">EXAMPLE: for-loops with column names</h1>

<p>When doing multiple nested for-loops it can be easier to keep track of the loop variable if is is named instead of just having a number. Use <b>--header :</b> to let the first argument be an named alias for the positional replacement string:</p>

<pre><code>  parallel --header : echo {gender} {size} ::: gender M F ::: size S M L XL XXL</code></pre>

<p>This also works if the input file is a file with columns:</p>

<pre><code>  cat addressbook.tsv | parallel --colsep &#39;\t&#39; --header : echo {Name} {E-mail address}</code></pre>

<h1 id="EXAMPLE:-Count-the-differences-between-all-files-in-a-dir">EXAMPLE: Count the differences between all files in a dir</h1>

<p>Using <b>--results</b> the results are saved in /tmp/diffcount*.</p>

<pre><code>  parallel --results /tmp/diffcount &quot;diff -U 0 {1} {2} |tail -n +3 |grep -v &#39;^@&#39;|wc -l&quot; ::: * ::: *</code></pre>

<p>To see the difference between file A and file B look at the file &#39;/tmp/diffcount/1/A/2/B&#39;.</p>

<h1 id="EXAMPLE:-Speeding-up-fast-jobs">EXAMPLE: Speeding up fast jobs</h1>

<p>Starting a job on the local machine takes around 10 ms. This can be a big overhead if the job takes very few ms to run. Often you can group small jobs together using <b>-X</b> which will make the overhead less significant. Compare the speed of these:</p>

<pre><code>    seq -w 0 9999 | parallel touch pict{}.jpg

    seq -w 0 9999 | parallel -X touch pict{}.jpg</code></pre>

<p>If your program cannot take multiple arguments, then you can use GNU <b>parallel</b> to spawn multiple GNU <b>parallel</b>s:</p>

<pre><code>    seq -w 0 999999 | parallel -j10 --pipe parallel -j0 touch pict{}.jpg</code></pre>

<p>If <b>-j0</b> normally spawns 252 jobs, then the above will try to spawn 2520 jobs. On a normal GNU/Linux system you can spawn 32000 jobs using this technique with no problems. To raise the 32000 jobs limit raise /proc/sys/kernel/pid_max to 4194303.</p>

<h1 id="EXAMPLE:-Using-shell-variables">EXAMPLE: Using shell variables</h1>

<p>When using shell variables you need to quote them correctly as they may otherwise be split on spaces.</p>

<p>Notice the difference between:</p>

<pre><code> V=(&quot;My brother&#39;s 12\&quot; records are worth &lt;\$\$\$&gt;&quot;&#39;!&#39; Foo Bar)
 parallel echo ::: ${V[@]} # This is probably not what you want</code></pre>

<p>and:</p>

<pre><code> V=(&quot;My brother&#39;s 12\&quot; records are worth &lt;\$\$\$&gt;&quot;&#39;!&#39; Foo Bar)
 parallel echo ::: &quot;${V[@]}&quot;</code></pre>

<p>When using variables in the actual command that contains special characters (e.g. space) you can quote them using <b>&#39;&quot;$VAR&quot;&#39;</b> or using &quot;&#39;s and <b>-q</b>:</p>

<pre><code> V=&quot;Here  are  two &quot;
 parallel echo &quot;&#39;$V&#39;&quot; ::: spaces
 parallel -q echo &quot;$V&quot; ::: spaces</code></pre>

<h1 id="EXAMPLE:-Group-output-lines">EXAMPLE: Group output lines</h1>

<p>When running jobs that output data, you often do not want the output of multiple jobs to run together. GNU <b>parallel</b> defaults to grouping the output of each job, so the output is printed when the job finishes. If you want full lines to be printed while the job is running you can use <b>--line-buffer</b>. If you want output to be printed as soon as possible you can use <b>-u</b>.</p>

<p>Compare the output of:</p>

<p><b>parallel traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<p>to the output of:</p>

<p><b>parallel --line-buffer traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<p>and:</p>

<p><b>parallel -u traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<h1 id="EXAMPLE:-Tag-output-lines">EXAMPLE: Tag output lines</h1>

<p>GNU <b>parallel</b> groups the output lines, but it can be hard to see where the different jobs begin. <b>--tag</b> prepends the argument to make that more visible:</p>

<p><b>parallel --tag traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<p><b>--tag</b> works with <b>--line-buffer</b> but not with <b>-u</b>:</p>

<p><b>parallel --tag --line-buffer traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<p>Check the uptime of the servers in <i>~/.parallel/sshloginfile</i>:</p>

<p><b>parallel --tag -S .. --nonall uptime</b></p>

<h1 id="EXAMPLE:-Keep-order-of-output-same-as-order-of-input">EXAMPLE: Keep order of output same as order of input</h1>

<p>Normally the output of a job will be printed as soon as it completes. Sometimes you want the order of the output to remain the same as the order of the input. This is often important, if the output is used as input for another system. <b>-k</b> will make sure the order of output will be in the same order as input even if later jobs end before earlier jobs.</p>

<p>Append a string to every line in a text file:</p>

<p><b>cat textfile | parallel -k echo {} append_string</b></p>

<p>If you remove <b>-k</b> some of the lines may come out in the wrong order.</p>

<p>Another example is <b>traceroute</b>:</p>

<p><b>parallel traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<p>will give traceroute of foss.org.my, debian.org and freenetproject.org, but it will be sorted according to which job completed first.</p>

<p>To keep the order the same as input run:</p>

<p><b>parallel -k traceroute ::: foss.org.my debian.org freenetproject.org</b></p>

<p>This will make sure the traceroute to foss.org.my will be printed first.</p>

<p>A bit more complex example is downloading a huge file in chunks in parallel: Some internet connections will deliver more data if you download files in parallel. For downloading files in parallel see: &quot;EXAMPLE: Download 10 images for each of the past 30 days&quot;. But if you are downloading a big file you can download the file in chunks in parallel.</p>

<p>To download byte 10000000-19999999 you can use <b>curl</b>:</p>

<p><b>curl -r 10000000-19999999 http://example.com/the/big/file</b> &gt; <b>file.part</b></p>

<p>To download a 1 GB file we need 100 10MB chunks downloaded and combined in the correct order.</p>

<p><b>seq 0 99 | parallel -k curl -r \ {}0000000-{}9999999 http://example.com/the/big/file</b> &gt; <b>file</b></p>

<h1 id="EXAMPLE:-Parallel-grep">EXAMPLE: Parallel grep</h1>

<p><b>grep -r</b> greps recursively through directories. On multicore CPUs GNU <b>parallel</b> can often speed this up.</p>

<p><b>find . -type f | parallel -k -j150% -n 1000 -m grep -H -n STRING {}</b></p>

<p>This will run 1.5 job per core, and give 1000 arguments to <b>grep</b>.</p>

<h1 id="EXAMPLE:-Grepping-n-lines-for-m-regular-expressions">EXAMPLE: Grepping n lines for m regular expressions.</h1>

<p>The simplest solution to grep a big file for a lot of regexps is:</p>

<pre><code>    grep -f regexps.txt bigfile</code></pre>

<p>Or if the regexps are fixed strings:</p>

<pre><code>    grep -F -f regexps.txt bigfile</code></pre>

<p>There are 2 limiting factors: CPU and disk I/O. CPU is easy to measure: If the grep takes &gt;90% CPU (e.g. when running top), then the CPU is a limiting factor, and parallelization will speed this up. If not, then disk I/O is the limiting factor, and depending on the disk system it may be faster or slower to parallelize. The only way to know for certain is to measure.</p>

<p>If the CPU is the limiting factor parallelization should be done on the regexps:</p>

<pre><code>    cat regexp.txt | parallel --pipe -L1000 --round-robin grep -f - bigfile</code></pre>

<p>If a line matches multiple regexps, the line may be duplicated. The command will start one grep per CPU and read bigfile one time per CPU, but as that is done in parallel, all reads except the first will be cached in RAM. Depending on the size of regexp.txt it may be faster to use --block 10m instead of -L1000. If regexp.txt is too big to fit in RAM, remove --round-robin and adjust -L1000. This will cause bigfile to be read more times.</p>

<p>Some storage systems perform better when reading multiple chunks in parallel. This is true for some RAID systems and for some network file systems. To parallelize the reading of bigfile:</p>

<pre><code>    parallel --pipepart --block 100M -a bigfile -k grep -f regexp.txt</code></pre>

<p>This will split bigfile into 100MB chunks and run grep on each of these chunks. To parallelize both reading of bigfile and regexp.txt combine the two using --fifo:</p>

<pre><code>    parallel --pipepart --block 100M -a bigfile --fifo cat regexp.txt \
    \| parallel --pipe -L1000 --round-robin grep -f - {}</code></pre>

<p>If a line matches multiple regexps, the line may be duplicated.</p>

<h1 id="EXAMPLE:-Using-remote-computers">EXAMPLE: Using remote computers</h1>

<p>To run commands on a remote computer SSH needs to be set up and you must be able to login without entering a password (The commands <b>ssh-copy-id</b> and <b>ssh-agent</b> may help you do that).</p>

<p>If you need to login to a whole cluster, you typically do not want to accept the host key for every host. You want to accept them the first time and be warned if they are ever changed. To do that:</p>

<pre><code>  # Add the servers to the sshloginfile
  (echo servera; echo serverb) &gt; .parallel/my_cluster
  # Make sure .ssh/config exist
  touch .ssh/config
  cp .ssh/config .ssh/config.backup
  # Disable StrictHostKeyChecking temporarily
  (echo &#39;Host *&#39;; echo StrictHostKeyChecking no) &gt;&gt; .ssh/config
  parallel --slf my_cluster --nonall true
  # Remove the disabling of StrictHostKeyChecking
  mv .ssh/config.backup .ssh/config</code></pre>

<p>The servers in <b>.parallel/my_cluster</b> are now added in <b>.ssh/known_hosts</b>.</p>

<p>To run <b>echo</b> on <b>server.example.com</b>:</p>

<pre><code>  seq 10 | parallel --sshlogin server.example.com echo</code></pre>

<p>To run commands on more than one remote computer run:</p>

<pre><code>  seq 10 | parallel --sshlogin server.example.com,server2.example.net echo</code></pre>

<p>Or:</p>

<pre><code>  seq 10 | parallel --sshlogin server.example.com \
    --sshlogin server2.example.net echo</code></pre>

<p>If the login username is <i>foo</i> on <i>server2.example.net</i> use:</p>

<pre><code>  seq 10 | parallel --sshlogin server.example.com \
    --sshlogin foo@server2.example.net echo</code></pre>

<p>If your list of hosts is <i>server1-88.example.net</i> with login <i>foo</i>:</p>

<pre><code>  seq 10 | parallel -Sfoo@server{1..88}.example.net echo</code></pre>

<p>To distribute the commands to a list of computers, make a file <i>mycomputers</i> with all the computers:</p>

<pre><code>  server.example.com
  foo@server2.example.com
  server3.example.com</code></pre>

<p>Then run:</p>

<pre><code>  seq 10 | parallel --sshloginfile mycomputers echo</code></pre>

<p>To include the local computer add the special sshlogin &#39;:&#39; to the list:</p>

<pre><code>  server.example.com
  foo@server2.example.com
  server3.example.com
  :</code></pre>

<p>GNU <b>parallel</b> will try to determine the number of CPU cores on each of the remote computers, and run one job per CPU core - even if the remote computers do not have the same number of CPU cores.</p>

<p>If the number of CPU cores on the remote computers is not identified correctly the number of CPU cores can be added in front. Here the computer has 8 CPU cores.</p>

<pre><code>  seq 10 | parallel --sshlogin 8/server.example.com echo</code></pre>

<h1 id="EXAMPLE:-Transferring-of-files">EXAMPLE: Transferring of files</h1>

<p>To recompress gzipped files with <b>bzip2</b> using a remote computer run:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | \
    parallel --sshlogin server.example.com \
    --transfer &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>This will list the .gz-files in the <i>logs</i> directory and all directories below. Then it will transfer the files to <i>server.example.com</i> to the corresponding directory in <i>$HOME/logs</i>. On <i>server.example.com</i> the file will be recompressed using <b>zcat</b> and <b>bzip2</b> resulting in the corresponding file with <i>.gz</i> replaced with <i>.bz2</i>.</p>

<p>If you want the resulting bz2-file to be transferred back to the local computer add <i>--return {.}.bz2</i>:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | \
    parallel --sshlogin server.example.com \
    --transfer --return {.}.bz2 &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>After the recompressing is done the <i>.bz2</i>-file is transferred back to the local computer and put next to the original <i>.gz</i>-file.</p>

<p>If you want to delete the transferred files on the remote computer add <i>--cleanup</i>. This will remove both the file transferred to the remote computer and the files transferred from the remote computer:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | \
    parallel --sshlogin server.example.com \
    --transfer --return {.}.bz2 --cleanup &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>If you want run on several computers add the computers to <i>--sshlogin</i> either using &#39;,&#39; or multiple <i>--sshlogin</i>:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | \
    parallel --sshlogin server.example.com,server2.example.com \
    --sshlogin server3.example.com \
    --transfer --return {.}.bz2 --cleanup &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>You can add the local computer using <i>--sshlogin :</i>. This will disable the removing and transferring for the local computer only:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | \
    parallel --sshlogin server.example.com,server2.example.com \
    --sshlogin server3.example.com \
    --sshlogin : \
    --transfer --return {.}.bz2 --cleanup &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>Often <i>--transfer</i>, <i>--return</i> and <i>--cleanup</i> are used together. They can be shortened to <i>--trc</i>:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | \
    parallel --sshlogin server.example.com,server2.example.com \
    --sshlogin server3.example.com \
    --sshlogin : \
    --trc {.}.bz2 &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>With the file <i>mycomputers</i> containing the list of computers it becomes:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | parallel --sshloginfile mycomputers \
    --trc {.}.bz2 &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<p>If the file <i>~/.parallel/sshloginfile</i> contains the list of computers the special short hand <i>-S ..</i> can be used:</p>

<pre><code>  find logs/ -name &#39;*.gz&#39; | parallel -S .. \
    --trc {.}.bz2 &quot;zcat {} | bzip2 -9 &gt;{.}.bz2&quot;</code></pre>

<h1 id="EXAMPLE:-Distributing-work-to-local-and-remote-computers">EXAMPLE: Distributing work to local and remote computers</h1>

<p>Convert *.mp3 to *.ogg running one process per CPU core on local computer and server2:</p>

<pre><code>  parallel --trc {.}.ogg -S server2,: \
  &#39;mpg321 -w - {} | oggenc -q0 - -o {.}.ogg&#39; ::: *.mp3</code></pre>

<h1 id="EXAMPLE:-Running-the-same-command-on-remote-computers">EXAMPLE: Running the same command on remote computers</h1>

<p>To run the command <b>uptime</b> on remote computers you can do:</p>

<p><b>parallel --tag --nonall -S server1,server2 uptime</b></p>

<p><b>--nonall</b> reads no arguments. If you have a list of jobs you want run on each computer you can do:</p>

<p><b>parallel --tag --onall -S server1,server2 echo ::: 1 2 3</b></p>

<p>Remove <b>--tag</b> if you do not want the sshlogin added before the output.</p>

<p>If you have a lot of hosts use &#39;-j0&#39; to access more hosts in parallel.</p>

<h1 id="EXAMPLE:-Parallelizing-rsync">EXAMPLE: Parallelizing rsync</h1>

<p><b>rsync</b> is a great tool, but sometimes it will not fill up the available bandwidth. This is often a problem when copying several big files over high speed connections.</p>

<p>The following will start one <b>rsync</b> per big file in <i>src-dir</i> to <i>dest-dir</i> on the server <i>fooserver</i>:</p>

<p><b>cd src-dir; find . -type f -size +100000 | parallel -v ssh fooserver mkdir -p /dest-dir/{//}\;rsync -s -Havessh {} fooserver:/dest-dir/{}</b></p>

<p>The dirs created may end up with wrong permissions and smaller files are not being transferred. To fix those run <b>rsync</b> a final time:</p>

<p><b>rsync -Havessh src-dir/ fooserver:/dest-dir/</b></p>

<p>If you are unable to push data, but need to pull them and the files are called digits.png (e.g. 000000.png) you might be able to do:</p>

<p><b>seq -w 0 99 | parallel rsync -Havessh fooserver:src-path/*{}.png destdir/</b></p>

<h1 id="EXAMPLE:-Use-multiple-inputs-in-one-command">EXAMPLE: Use multiple inputs in one command</h1>

<p>Copy files like foo.es.ext to foo.ext:</p>

<p><b>ls *.es.* | perl -pe &#39;print; s/\.es//&#39; | parallel -N2 cp {1} {2}</b></p>

<p>The perl command spits out 2 lines for each input. GNU <b>parallel</b> takes 2 inputs (using <b>-N2</b>) and replaces {1} and {2} with the inputs.</p>

<p>Count in binary:</p>

<p><b>parallel -k echo ::: 0 1 ::: 0 1 ::: 0 1 ::: 0 1 ::: 0 1 ::: 0 1</b></p>

<p>Print the number on the opposing sides of a six sided die:</p>

<p><b>parallel --xapply -a &lt;(seq 6) -a &lt;(seq 6 -1 1) echo</b></p>

<p><b>parallel --xapply echo :::: &lt;(seq 6) &lt;(seq 6 -1 1)</b></p>

<p>Convert files from all subdirs to PNG-files with consecutive numbers (useful for making input PNG&#39;s for <b>ffmpeg</b>):</p>

<p><b>parallel --xapply -a &lt;(find . -type f | sort) -a &lt;(seq $(find . -type f|wc -l)) convert {1} {2}.png</b></p>

<p>Alternative version:</p>

<p><b>find . -type f | sort | parallel convert {} {#}.png</b></p>

<h1 id="EXAMPLE:-Use-a-table-as-input">EXAMPLE: Use a table as input</h1>

<p>Content of table_file.tsv:</p>

<pre><code>  foo&lt;TAB&gt;bar
  baz &lt;TAB&gt; quux</code></pre>

<p>To run:</p>

<pre><code>  cmd -o bar -i foo
  cmd -o quux -i baz</code></pre>

<p>you can run:</p>

<p><b>parallel -a table_file.tsv --colsep &#39;\t&#39; cmd -o {2} -i {1}</b></p>

<p>Note: The default for GNU <b>parallel</b> is to remove the spaces around the columns. To keep the spaces:</p>

<p><b>parallel -a table_file.tsv --trim n --colsep &#39;\t&#39; cmd -o {2} -i {1}</b></p>

<h1 id="EXAMPLE:-Run-the-same-command-10-times">EXAMPLE: Run the same command 10 times</h1>

<p>If you want to run the same command with the same arguments 10 times in parallel you can do:</p>

<p><b>seq 10 | parallel -n0 my_command my_args</b></p>

<h1 id="EXAMPLE:-Working-as-cat-sh.-Resource-inexpensive-jobs-and-evaluation">EXAMPLE: Working as cat | sh. Resource inexpensive jobs and evaluation</h1>

<p>GNU <b>parallel</b> can work similar to <b>cat | sh</b>.</p>

<p>A resource inexpensive job is a job that takes very little CPU, disk I/O and network I/O. Ping is an example of a resource inexpensive job. wget is too - if the webpages are small.</p>

<p>The content of the file jobs_to_run:</p>

<pre><code>  ping -c 1 10.0.0.1
  wget http://example.com/status.cgi?ip=10.0.0.1
  ping -c 1 10.0.0.2
  wget http://example.com/status.cgi?ip=10.0.0.2
  ...
  ping -c 1 10.0.0.255
  wget http://example.com/status.cgi?ip=10.0.0.255</code></pre>

<p>To run 100 processes simultaneously do:</p>

<p><b>parallel -j 100 &lt; jobs_to_run</b></p>

<p>As there is not a <i>command</i> the jobs will be evaluated by the shell.</p>

<h1 id="EXAMPLE:-Processing-a-big-file-using-more-cores">EXAMPLE: Processing a big file using more cores</h1>

<p>To process a big file or some output you can use <b>--pipe</b> to split up the data into blocks and pipe the blocks into the processing program.</p>

<p>If the program is <b>gzip -9</b> you can do:</p>

<p><b>cat bigfile | parallel --pipe --recend &#39;&#39; -k gzip -9 </b>&gt;<b>bigfile.gz</b></p>

<p>This will split <b>bigfile</b> into blocks of 1 MB and pass that to <b>gzip -9</b> in parallel. One <b>gzip</b> will be run per CPU core. The output of <b>gzip -9</b> will be kept in order and saved to <b>bigfile.gz</b></p>

<p><b>gzip</b> works fine if the output is appended, but some processing does not work like that - for example sorting. For this GNU <b>parallel</b> can put the output of each command into a file. This will sort a big file in parallel:</p>

<p><b>cat bigfile | parallel --pipe --files sort | parallel -Xj1 sort -m {} &#39;;&#39; rm {} </b>&gt;<b>bigfile.sort</b></p>

<p>Here <b>bigfile</b> is split into blocks of around 1MB, each block ending in &#39;\n&#39; (which is the default for <b>--recend</b>). Each block is passed to <b>sort</b> and the output from <b>sort</b> is saved into files. These files are passed to the second <b>parallel</b> that runs <b>sort -m</b> on the files before it removes the files. The output is saved to <b>bigfile.sort</b>.</p>

<p>GNU <b>parallel</b>&#39;s <b>--pipe</b> maxes out at around 100 MB/s because every byte has to be copied through GNU <b>parallel</b>. But if <b>bigfile</b> is a real (seekable) file GNU <b>parallel</b> can by-pass the copying and send the parts directly to the program:</p>

<p><b>parallel --pipepart --block 100m -a bigfile --files sort | parallel -Xj1 sort -m {} &#39;;&#39; rm {} </b>&gt;<b>bigfile.sort</b></p>

<h1 id="EXAMPLE:-Running-more-than-250-jobs-workaround">EXAMPLE: Running more than 250 jobs workaround</h1>

<p>If you need to run a massive amount of jobs in parallel, then you will likely hit the filehandle limit which is often around 250 jobs. If you are super user you can raise the limit in /etc/security/limits.conf but you can also use this workaround. The filehandle limit is per process. That means that if you just spawn more GNU <b>parallel</b>s then each of them can run 250 jobs. This will spawn up to 2500 jobs:</p>

<p><b>cat myinput | parallel --pipe -N 50 --round-robin -j50 parallel -j50 your_prg</b></p>

<p>This will spawn up to 62500 jobs (use with caution - you need 64 GB RAM to do this, and you may need to increase /proc/sys/kernel/pid_max):</p>

<p><b>cat myinput | parallel --pipe -N 250 --round-robin -j250 parallel -j250 your_prg</b></p>

<h1 id="EXAMPLE:-Working-as-mutex-and-counting-semaphore">EXAMPLE: Working as mutex and counting semaphore</h1>

<p>The command <b>sem</b> is an alias for <b>parallel --semaphore</b>.</p>

<p>A counting semaphore will allow a given number of jobs to be started in the background. When the number of jobs are running in the background, GNU <b>sem</b> will wait for one of these to complete before starting another command. <b>sem --wait</b> will wait for all jobs to complete.</p>

<p>Run 10 jobs concurrently in the background:</p>

<pre><code>  for i in *.log ; do
    echo $i
    sem -j10 gzip $i &quot;;&quot; echo done
  done
  sem --wait</code></pre>

<p>A mutex is a counting semaphore allowing only one job to run. This will edit the file <i>myfile</i> and prepends the file with lines with the numbers 1 to 3.</p>

<pre><code>  seq 3 | parallel sem sed -i -e &#39;i{}&#39; myfile</code></pre>

<p>As <i>myfile</i> can be very big it is important only one process edits the file at the same time.</p>

<p>Name the semaphore to have multiple different semaphores active at the same time:</p>

<pre><code>  seq 3 | parallel sem --id mymutex sed -i -e &#39;i{}&#39; myfile</code></pre>

<h1 id="EXAMPLE:-Start-editor-with-filenames-from-stdin-standard-input">EXAMPLE: Start editor with filenames from stdin (standard input)</h1>

<p>You can use GNU <b>parallel</b> to start interactive programs like emacs or vi:</p>

<p><b>cat filelist | parallel --tty -X emacs</b></p>

<p><b>cat filelist | parallel --tty -X vi</b></p>

<p>If there are more files than will fit on a single command line, the editor will be started again with the remaining files.</p>

<h1 id="EXAMPLE:-Running-sudo">EXAMPLE: Running sudo</h1>

<p><b>sudo</b> requires a password to run a command as root. It caches the access, so you only need to enter the password again if you have not used <b>sudo</b> for a while.</p>

<p>The command:</p>

<pre><code>  parallel sudo echo ::: This is a bad idea</code></pre>

<p>is no good, as you would be prompted for the sudo password for each of the jobs. You can either do:</p>

<pre><code>  sudo echo This
  parallel sudo echo ::: is a good idea</code></pre>

<p>or:</p>

<pre><code>  sudo parallel echo ::: This is a good idea</code></pre>

<p>This way you only have to enter the sudo password once.</p>

<h1 id="EXAMPLE:-GNU-Parallel-as-queue-system-batch-manager">EXAMPLE: GNU Parallel as queue system/batch manager</h1>

<p>GNU <b>parallel</b> can work as a simple job queue system or batch manager. The idea is to put the jobs into a file and have GNU <b>parallel</b> read from that continuously. As GNU <b>parallel</b> will stop at end of file we use <b>tail</b> to continue reading:</p>

<p><b>true </b>&gt;<b>jobqueue</b>; <b>tail -n+0 -f jobqueue | parallel</b></p>

<p>To submit your jobs to the queue:</p>

<p><b>echo my_command my_arg </b>&gt;&gt;<b> jobqueue</b></p>

<p>You can of course use <b>-S</b> to distribute the jobs to remote computers:</p>

<p><b>true </b>&gt;<b>jobqueue</b>; <b>tail -f jobqueue | parallel -S ..</b></p>

<p>There is a a small issue when using GNU <b>parallel</b> as queue system/batch manager: You have to submit JobSlot number of jobs before they will start, and after that you can submit one at a time, and job will start immediately if free slots are available. Output from the running or completed jobs are held back and will only be printed when JobSlots more jobs has been started (unless you use --ungroup or -u, in which case the output from the jobs are printed immediately). E.g. if you have 10 jobslots then the output from the first completed job will only be printed when job 11 has started, and the output of second completed job will only be printed when job 12 has started.</p>

<p>To use <b>--eof</b> to make GNU <b>parallel</b> exit, <b>tail</b> also needs to be forced to exit:</p>

<pre><code>    tail -n+0 -f command-list.txt |
      (parallel --eof=EXIT {}; echo Parallel is now done;
       (seq 1000 &gt;&gt; command-list.txt &amp;);
      echo Done appending dummy data forcing tail to exit)</code></pre>

<h1 id="EXAMPLE:-GNU-Parallel-as-dir-processor">EXAMPLE: GNU Parallel as dir processor</h1>

<p>If you have a dir in which users drop files that needs to be processed you can do this on GNU/Linux (If you know what <b>inotifywait</b> is called on other platforms file a bug report):</p>

<p><b>inotifywait -q -m -r -e MOVED_TO -e CLOSE_WRITE --format %w%f my_dir | parallel -u echo</b></p>

<p>This will run the command <b>echo</b> on each file put into <b>my_dir</b> or subdirs of <b>my_dir</b>.</p>

<p>You can of course use <b>-S</b> to distribute the jobs to remote computers:</p>

<p><b>inotifywait -q -m -r -e MOVED_TO -e CLOSE_WRITE --format %w%f my_dir | parallel -S .. -u echo</b></p>

<p>If the files to be processed are in a tar file then unpacking one file and processing it immediately may be faster than first unpacking all files. Set up the dir processor as above and unpack into the dir.</p>

<p>Using GNU Parallel as dir processor has the same limitations as using GNU Parallel as queue system/batch manager.</p>

<h1 id="QUOTING">QUOTING</h1>

<p>GNU <b>parallel</b> is very liberal in quoting. You only need to quote characters that have special meaning in shell:</p>

<p>( ) $ ` &#39; &quot; &lt; &gt; ; | \</p>

<p>and depending on context these needs to be quoted, too:</p>

<p>~ &amp; # ! ? space * {</p>

<p>Therefore most people will never need more quoting than putting &#39;\&#39; in front of the special characters.</p>

<p>Often you can simply put \&#39; around every &#39;:</p>

<pre><code>  perl -ne &#39;/^\S+\s+\S+$/ and print $ARGV,&quot;\n&quot;&#39; file</code></pre>

<p>can be quoted:</p>

<pre><code>  parallel perl -ne \&#39;&#39;/^\S+\s+\S+$/ and print $ARGV,&quot;\n&quot;&#39;\&#39; ::: file</code></pre>

<p>However, when you want to use a shell variable you need to quote the $-sign. Here is an example using $PARALLEL_SEQ. This variable is set by GNU <b>parallel</b> itself, so the evaluation of the $ must be done by the sub shell started by GNU <b>parallel</b>:</p>

<p><b>seq 10 | parallel -N2 echo seq:\$PARALLEL_SEQ arg1:{1} arg2:{2}</b></p>

<p>If the variable is set before GNU <b>parallel</b> starts you can do this:</p>

<p><b>VAR=this_is_set_before_starting</b></p>

<p><b>echo test | parallel echo {} $VAR</b></p>

<p>Prints: <b>test this_is_set_before_starting</b></p>

<p>It is a little more tricky if the variable contains more than one space in a row:</p>

<p><b>VAR=&quot;two spaces between each word&quot;</b></p>

<p><b>echo test | parallel echo {} \&#39;&quot;$VAR&quot;\&#39;</b></p>

<p>Prints: <b>test two spaces between each word</b></p>

<p>If the variable should not be evaluated by the shell starting GNU <b>parallel</b> but be evaluated by the sub shell started by GNU <b>parallel</b>, then you need to quote it:</p>

<p><b>echo test | parallel VAR=this_is_set_after_starting \; echo {} \$VAR</b></p>

<p>Prints: <b>test this_is_set_after_starting</b></p>

<p>It is a little more tricky if the variable contains space:</p>

<p><b>echo test | parallel VAR=&#39;&quot;two spaces between each word&quot;&#39; echo {} \&#39;&quot;$VAR&quot;\&#39;</b></p>

<p>Prints: <b>test two spaces between each word</b></p>

<p>$$ is the shell variable containing the process id of the shell. This will print the process id of the shell running GNU <b>parallel</b>:</p>

<p><b>seq 10 | parallel echo $$</b></p>

<p>And this will print the process ids of the sub shells started by GNU <b>parallel</b>.</p>

<p><b>seq 10 | parallel echo \$\$</b></p>

<p>If the special characters should not be evaluated by the sub shell then you need to protect it against evaluation from both the shell starting GNU <b>parallel</b> and the sub shell:</p>

<p><b>echo test | parallel echo {} \\\$VAR</b></p>

<p>Prints: <b>test $VAR</b></p>

<p>GNU <b>parallel</b> can protect against evaluation by the sub shell by using -q:</p>

<p><b>echo test | parallel -q echo {} \$VAR</b></p>

<p>Prints: <b>test $VAR</b></p>

<p>This is particularly useful if you have lots of quoting. If you want to run a perl script like this:</p>

<p><b>perl -ne &#39;/^\S+\s+\S+$/ and print $ARGV,&quot;\n&quot;&#39; file</b></p>

<p>It needs to be quoted like this:</p>

<p><b>ls | parallel perl -ne &#39;/^\\S+\\s+\\S+\$/\ and\ print\ \$ARGV,\&quot;\\n\&quot;&#39;</b> <b>ls | parallel perl -ne \&#39;&#39;/^\S+\s+\S+$/ and print $ARGV,&quot;\n&quot;&#39;\&#39;</b></p>

<p>Notice how spaces, \&#39;s, &quot;&#39;s, and $&#39;s need to be quoted. GNU <b>parallel</b> can do the quoting by using option -q:</p>

<p><b>ls | parallel -q perl -ne &#39;/^\S+\s+\S+$/ and print $ARGV,&quot;\n&quot;&#39;</b></p>

<p>However, this means you cannot make the sub shell interpret special characters. For example because of <b>-q</b> this WILL NOT WORK:</p>

<p><b>ls *.gz | parallel -q &quot;zcat {} </b>&gt;<b>{.}&quot;</b></p>

<p><b>ls *.gz | parallel -q &quot;zcat {} | bzip2 </b>&gt;<b>{.}.bz2&quot;</b></p>

<p>because &gt; and | need to be interpreted by the sub shell.</p>

<p>If you get errors like:</p>

<pre><code>  sh: -c: line 0: syntax error near unexpected token
  sh: Syntax error: Unterminated quoted string
  sh: -c: line 0: unexpected EOF while looking for matching `&#39;&#39;
  sh: -c: line 1: syntax error: unexpected end of file</code></pre>

<p>then you might try using <b>-q</b>.</p>

<p>If you are using <b>bash</b> process substitution like <b>&lt;(cat foo)</b> then you may try <b>-q</b> and prepending <i>command</i> with <b>bash -c</b>:</p>

<p><b>ls | parallel -q bash -c &#39;wc -c &lt;(echo {})&#39;</b></p>

<p>Or for substituting output:</p>

<p><b>ls | parallel -q bash -c &#39;tar c {} | tee </b>&gt;<b>(gzip </b>&gt;<b>{}.tar.gz) | bzip2 </b>&gt;<b>{}.tar.bz2&#39;</b></p>

<p><b>Conclusion</b>: To avoid dealing with the quoting problems it may be easier just to write a small script or a function (remember to <b>export -f</b> the function) and have GNU <b>parallel</b> call that.</p>

<h1 id="LIST-RUNNING-JOBS">LIST RUNNING JOBS</h1>

<p>If you want a list of the jobs currently running you can run:</p>

<p><b>killall -USR1 parallel</b></p>

<p>GNU <b>parallel</b> will then print the currently running jobs on stderr (standard error).</p>

<h1 id="COMPLETE-RUNNING-JOBS-BUT-DO-NOT-START-NEW-JOBS">COMPLETE RUNNING JOBS BUT DO NOT START NEW JOBS</h1>

<p>If you regret starting a lot of jobs you can simply break GNU <b>parallel</b>, but if you want to make sure you do not have half-completed jobs you should send the signal <b>SIGTERM</b> to GNU <b>parallel</b>:</p>

<p><b>killall -TERM parallel</b></p>

<p>This will tell GNU <b>parallel</b> to not start any new jobs, but wait until the currently running jobs are finished before exiting.</p>

<h1 id="ENVIRONMENT-VARIABLES">ENVIRONMENT VARIABLES</h1>

<dl>

<dt id="PARALLEL_PID">$PARALLEL_PID</dt>
<dd>

<p>The environment variable $PARALLEL_PID is set by GNU <b>parallel</b> and is visible to the jobs started from GNU <b>parallel</b>. This makes it possible for the jobs to communicate directly to GNU <b>parallel</b>. Remember to quote the $, so it gets evaluated by the correct shell.</p>

<p><b>Example:</b> If each of the jobs tests a solution and one of jobs finds the solution the job can tell GNU <b>parallel</b> not to start more jobs by: <b>kill -TERM $PARALLEL_PID</b>. This only works on the local computer.</p>

</dd>
<dt id="PARALLEL_SHELL-alpha-testing">$PARALLEL_SHELL (alpha testing)</dt>
<dd>

<p>Use this shell the shell for the commands run by GNU Parallel:</p>

<ul>

<li><p>$PARALLEL_SHELL. If undefined use:</p>

</li>
<li><p>The shell that started GNU Parallel. If that cannot be determined:</p>

</li>
<li><p>$SHELL. If undefined use:</p>

</li>
<li><p>/bin/sh</p>

</li>
</ul>

</dd>
<dt id="PARALLEL_SEQ">$PARALLEL_SEQ</dt>
<dd>

<p>$PARALLEL_SEQ will be set to the sequence number of the job running. Remember to quote the $, so it gets evaluated by the correct shell.</p>

<p><b>Example:</b></p>

<p><b>seq 10 | parallel -N2 echo seq:&#39;$&#39;PARALLEL_SEQ arg1:{1} arg2:{2}</b></p>

</dd>
<dt id="TMPDIR">$TMPDIR</dt>
<dd>

<p>Directory for temporary files. See: <b>--tmpdir</b>.</p>

</dd>
<dt id="PARALLEL">$PARALLEL</dt>
<dd>

<p>The environment variable $PARALLEL will be used as default options for GNU <b>parallel</b>. If the variable contains special shell characters (e.g. $, *, or space) then these need to be to be escaped with \.</p>

<p><b>Example:</b></p>

<p><b>cat list | parallel -j1 -k -v ls</b></p>

<p>can be written as:</p>

<p><b>cat list | PARALLEL=&quot;-kvj1&quot; parallel ls</b></p>

<p><b>cat list | parallel -j1 -k -v -S&quot;myssh user@server&quot; ls</b></p>

<p>can be written as:</p>

<p><b>cat list | PARALLEL=&#39;-kvj1 -S myssh\ user@server&#39; parallel echo</b></p>

<p>Notice the \ in the middle is needed because &#39;myssh&#39; and &#39;user@server&#39; must be one argument.</p>

</dd>
</dl>

<h1 id="DEFAULT-PROFILE-CONFIG-FILE">DEFAULT PROFILE (CONFIG FILE)</h1>

<p>The global configuration file /etc/parallel/config, followed by user configuration file ~/.parallel/config (formerly known as .parallelrc) will be read in turn if they exist. Lines starting with &#39;#&#39; will be ignored. The format can follow that of the environment variable $PARALLEL, but it is often easier to simply put each option on its own line.</p>

<p>Options on the command line take precedence, followed by the environment variable $PARALLEL, user configuration file ~/.parallel/config, and finally the global configuration file /etc/parallel/config.</p>

<p>Note that no file that is read for options, nor the environment variable $PARALLEL, may contain retired options such as <b>--tollef</b>.</p>

<h1 id="PROFILE-FILES">PROFILE FILES</h1>

<p>If <b>--profile</b> set, GNU <b>parallel</b> will read the profile from that file rather than the global or user configuration files. You can have multiple <b>--profiles</b>.</p>

<p>Example: Profile for running a command on every sshlogin in ~/.ssh/sshlogins and prepend the output with the sshlogin:</p>

<pre><code>  echo --tag -S .. --nonall &gt; ~/.parallel/n
  parallel -Jn uptime</code></pre>

<p>Example: Profile for running every command with <b>-j-1</b> and <b>nice</b></p>

<pre><code>  echo -j-1 nice &gt; ~/.parallel/nice_profile
  parallel -J nice_profile bzip2 -9 ::: *</code></pre>

<p>Example: Profile for running a perl script before every command:</p>

<pre><code>  echo &quot;perl -e &#39;\$a=\$\$; print \$a,\&quot; \&quot;,&#39;\$PARALLEL_SEQ&#39;,\&quot; \&quot;;&#39;;&quot; &gt; ~/.parallel/pre_perl
  parallel -J pre_perl echo ::: *</code></pre>

<p>Note how the $ and &quot; need to be quoted using \.</p>

<p>Example: Profile for running distributed jobs with <b>nice</b> on the remote computers:</p>

<pre><code>  echo -S .. nice &gt; ~/.parallel/dist
  parallel -J dist --trc {.}.bz2 bzip2 -9 ::: *</code></pre>

<h1 id="EXIT-STATUS">EXIT STATUS</h1>

<p>If <b>--halt-on-error</b> 0 or not specified:</p>

<dl>

<dt id="pod01">0</dt>
<dd>

<p>All jobs ran without error.</p>

</dd>
<dt id="pod1-253">1-253</dt>
<dd>

<p>Some of the jobs failed. The exit status gives the number of failed jobs</p>

</dd>
<dt id="pod254">254</dt>
<dd>

<p>More than 253 jobs failed.</p>

</dd>
<dt id="pod255">255</dt>
<dd>

<p>Other error.</p>

</dd>
</dl>

<p>If <b>--halt-on-error</b> 1 or 2: Exit status of the failing job.</p>

<h1 id="DIFFERENCES-BETWEEN-GNU-Parallel-AND-ALTERNATIVES">DIFFERENCES BETWEEN GNU Parallel AND ALTERNATIVES</h1>

<p>There are a lot programs with some of the functionality of GNU <b>parallel</b>. GNU <b>parallel</b> strives to include the best of the functionality without sacrificing ease of use.</p>

<h2 id="SUMMARY-TABLE">SUMMARY TABLE</h2>

<p>The following features are in some of the comparable tools:</p>

<p>Inputs I1. Arguments can be read from stdin I2. Arguments can be read from a file I3. Arguments can be read from multiple files I4. Arguments can be read from command line I5. Arguments can be read from a table I6. Arguments can be read from the same file using #! (shebang) I7. Line oriented input as default (Quoting of special chars not needed)</p>

<p>Manipulation of input M1. Composed command M2. Multiple arguments can fill up an execution line M3. Arguments can be put anywhere in the execution line M4. Multiple arguments can be put anywhere in the execution line M5. Arguments can be replaced with context M6. Input can be treated as complete execution line</p>

<p>Outputs O1. Grouping output so output from different jobs do not mix O2. Send stderr (standard error) to stderr (standard error) O3. Send stdout (standard output) to stdout (standard output) O4. Order of output can be same as order of input O5. Stdout only contains stdout (standard output) from the command O6. Stderr only contains stderr (standard error) from the command</p>

<p>Execution E1. Running jobs in parallel E2. List running jobs E3. Finish running jobs, but do not start new jobs E4. Number of running jobs can depend on number of cpus E5. Finish running jobs, but do not start new jobs after first failure E6. Number of running jobs can be adjusted while running</p>

<p>Remote execution R1. Jobs can be run on remote computers R2. Basefiles can be transferred R3. Argument files can be transferred R4. Result files can be transferred R5. Cleanup of transferred files R6. No config files needed R7. Do not run more than SSHD&#39;s MaxStartups can handle R8. Configurable SSH command R9. Retry if connection breaks occasionally</p>

<p>Semaphore S1. Possibility to work as a mutex S2. Possibility to work as a counting semaphore</p>

<p>Legend - = no x = not applicable ID = yes</p>

<p>As every new version of the programs are not tested the table may be outdated. Please file a bug-report if you find errors (See REPORTING BUGS).</p>

<p>parallel: I1 I2 I3 I4 I5 I6 I7 M1 M2 M3 M4 M5 M6 O1 O2 O3 O4 O5 O6 E1 E2 E3 E4 E5 E6 R1 R2 R3 R4 R5 R6 R7 R8 R9 S1 S2</p>

<p>xargs: I1 I2 - - - - - - M2 M3 - - - - O2 O3 - O5 O6 E1 - - - - - - - - - - x - - - - -</p>

<p>find -exec: - - - x - x - - M2 M3 - - - - - O2 O3 O4 O5 O6 - - - - - - - - - - - - - - - - x x</p>

<p>make -j: - - - - - - - - - - - - - O1 O2 O3 - x O6 E1 - - - E5 - - - - - - - - - - - -</p>

<p>ppss: I1 I2 - - - - I7 M1 - M3 - - M6 O1 - - x - - E1 E2 ?E3 E4 - - R1 R2 R3 R4 - - ?R7 ? ? - -</p>

<p>pexec: I1 I2 - I4 I5 - - M1 - M3 - - M6 O1 O2 O3 - O5 O6 E1 - - E4 - E6 R1 - - - - R6 - - - S1 -</p>

<p>xjobs: TODO - Please file a bug-report if you know what features xjobs supports (See REPORTING BUGS).</p>

<p>prll: TODO - Please file a bug-report if you know what features prll supports (See REPORTING BUGS).</p>

<p>dxargs: TODO - Please file a bug-report if you know what features dxargs supports (See REPORTING BUGS).</p>

<p>mdm/middelman: TODO - Please file a bug-report if you know what features mdm/middelman supports (See REPORTING BUGS).</p>

<p>xapply: TODO - Please file a bug-report if you know what features xapply supports (See REPORTING BUGS).</p>

<p>paexec: TODO - Please file a bug-report if you know what features paexec supports (See REPORTING BUGS).</p>

<p>ladon: TODO - Please file a bug-report if you know what features ladon supports (See REPORTING BUGS).</p>

<p>ClusterSSH: TODO - Please file a bug-report if you know what features ClusterSSH supports (See REPORTING BUGS).</p>

<h2 id="DIFFERENCES-BETWEEN-xargs-AND-GNU-Parallel">DIFFERENCES BETWEEN xargs AND GNU Parallel</h2>

<p><b>xargs</b> offers some of the same possibilities as GNU <b>parallel</b>.</p>

<p><b>xargs</b> deals badly with special characters (such as space, &#39; and &quot;). To see the problem try this:</p>

<pre><code>  touch important_file
  touch &#39;not important_file&#39;
  ls not* | xargs rm
  mkdir -p &quot;My brother&#39;s 12\&quot; records&quot;
  ls | xargs rmdir</code></pre>

<p>You can specify <b>-0</b> or <b>-d &quot;\n&quot;</b>, but many input generators are not optimized for using <b>NUL</b> as separator but are optimized for <b>newline</b> as separator. E.g <b>head</b>, <b>tail</b>, <b>awk</b>, <b>ls</b>, <b>echo</b>, <b>sed</b>, <b>tar -v</b>, <b>perl</b> (<b>-0</b> and \0 instead of \n), <b>locate</b> (requires using <b>-0</b>), <b>find</b> (requires using <b>-print0</b>), <b>grep</b> (requires user to use <b>-z</b> or <b>-Z</b>), <b>sort</b> (requires using <b>-z</b>).</p>

<p>So GNU <b>parallel</b>&#39;s newline separation can be emulated with:</p>

<p><b>cat | xargs -d &quot;\n&quot; -n1 <i>command</i></b></p>

<p><b>xargs</b> can run a given number of jobs in parallel, but has no support for running number-of-cpu-cores jobs in parallel.</p>

<p><b>xargs</b> has no support for grouping the output, therefore output may run together, e.g. the first half of a line is from one process and the last half of the line is from another process. The example <b>Parallel grep</b> cannot be done reliably with <b>xargs</b> because of this. To see this in action try:</p>

<pre><code>  parallel perl -e &#39;\$a=\&quot;1{}\&quot;x10000000\;print\ \$a,\&quot;\\n\&quot;&#39; &#39;&gt;&#39; {} ::: a b c d e f
  ls -l a b c d e f
  parallel -kP4 -n1 grep 1 &gt; out.par ::: a b c d e f
  echo a b c d e f | xargs -P4 -n1 grep 1 &gt; out.xargs-unbuf
  echo a b c d e f | xargs -P4 -n1 grep --line-buffered 1 &gt; out.xargs-linebuf
  echo a b c d e f | xargs -n1 grep 1 &gt; out.xargs-serial
  ls -l out*
  md5sum out*</code></pre>

<p><b>xargs</b> has no support for keeping the order of the output, therefore if running jobs in parallel using <b>xargs</b> the output of the second job cannot be postponed till the first job is done.</p>

<p><b>xargs</b> has no support for running jobs on remote computers.</p>

<p><b>xargs</b> has no support for context replace, so you will have to create the arguments.</p>

<p>If you use a replace string in <b>xargs</b> (<b>-I</b>) you can not force <b>xargs</b> to use more than one argument.</p>

<p>Quoting in <b>xargs</b> works like <b>-q</b> in GNU <b>parallel</b>. This means composed commands and redirection require using <b>bash -c</b>.</p>

<p><b>ls | parallel &quot;wc {} </b>&gt; <b>{}.wc&quot;</b></p>

<p>becomes (assuming you have 8 cores)</p>

<p><b>ls | xargs -d &quot;\n&quot; -P8 -I {} bash -c &quot;wc {} </b>&gt;<b> {}.wc&quot;</b></p>

<p>and</p>

<p><b>ls | parallel &quot;echo {}; ls {}|wc&quot;</b></p>

<p>becomes (assuming you have 8 cores)</p>

<p><b>ls | xargs -d &quot;\n&quot; -P8 -I {} bash -c &quot;echo {}; ls {}|wc&quot;</b></p>

<h2 id="DIFFERENCES-BETWEEN-find--exec-AND-GNU-Parallel">DIFFERENCES BETWEEN find -exec AND GNU Parallel</h2>

<p><b>find -exec</b> offer some of the same possibilities as GNU <b>parallel</b>.</p>

<p><b>find -exec</b> only works on files. So processing other input (such as hosts or URLs) will require creating these inputs as files. <b>find -exec</b> has no support for running commands in parallel.</p>

<h2 id="DIFFERENCES-BETWEEN-make--j-AND-GNU-Parallel">DIFFERENCES BETWEEN make -j AND GNU Parallel</h2>

<p><b>make -j</b> can run jobs in parallel, but requires a crafted Makefile to do this. That results in extra quoting to get filename containing newline to work correctly.</p>

<p><b>make -j</b> has no support for grouping the output, therefore output may run together, e.g. the first half of a line is from one process and the last half of the line is from another process. The example <b>Parallel grep</b> cannot be done reliably with <b>make -j</b> because of this.</p>

<p>(Very early versions of GNU <b>parallel</b> were coincidently implemented using <b>make -j</b>).</p>

<h2 id="DIFFERENCES-BETWEEN-ppss-AND-GNU-Parallel">DIFFERENCES BETWEEN ppss AND GNU Parallel</h2>

<p><b>ppss</b> is also a tool for running jobs in parallel.</p>

<p>The output of <b>ppss</b> is status information and thus not useful for using as input for another command. The output from the jobs are put into files.</p>

<p>The argument replace string ($ITEM) cannot be changed. Arguments must be quoted - thus arguments containing special characters (space &#39;&quot;&amp;!*) may cause problems. More than one argument is not supported. File names containing newlines are not processed correctly. When reading input from a file null cannot be used as a terminator. <b>ppss</b> needs to read the whole input file before starting any jobs.</p>

<p>Output and status information is stored in ppss_dir and thus requires cleanup when completed. If the dir is not removed before running <b>ppss</b> again it may cause nothing to happen as <b>ppss</b> thinks the task is already done. GNU <b>parallel</b> will normally not need cleaning up if running locally and will only need cleaning up if stopped abnormally and running remote (<b>--cleanup</b> may not complete if stopped abnormally). The example <b>Parallel grep</b> would require extra postprocessing if written using <b>ppss</b>.</p>

<p>For remote systems PPSS requires 3 steps: config, deploy, and start. GNU <b>parallel</b> only requires one step.</p>

<h3 id="EXAMPLES-FROM-ppss-MANUAL">EXAMPLES FROM ppss MANUAL</h3>

<p>Here are the examples from <b>ppss</b>&#39;s manual page with the equivalent using GNU <b>parallel</b>:</p>

<p><b>1</b> ./ppss.sh standalone -d /path/to/files -c &#39;gzip &#39;</p>

<p><b>1</b> find /path/to/files -type f | parallel gzip</p>

<p><b>2</b> ./ppss.sh standalone -d /path/to/files -c &#39;cp &quot;$ITEM&quot; /destination/dir &#39;</p>

<p><b>2</b> find /path/to/files -type f | parallel cp {} /destination/dir</p>

<p><b>3</b> ./ppss.sh standalone -f list-of-urls.txt -c &#39;wget -q &#39;</p>

<p><b>3</b> parallel -a list-of-urls.txt wget -q</p>

<p><b>4</b> ./ppss.sh standalone -f list-of-urls.txt -c &#39;wget -q &quot;$ITEM&quot;&#39;</p>

<p><b>4</b> parallel -a list-of-urls.txt wget -q {}</p>

<p><b>5</b> ./ppss config -C config.cfg -c &#39;encode.sh &#39; -d /source/dir -m 192.168.1.100 -u ppss -k ppss-key.key -S ./encode.sh -n nodes.txt -o /some/output/dir --upload --download ; ./ppss deploy -C config.cfg ; ./ppss start -C config</p>

<p><b>5</b> # parallel does not use configs. If you want a different username put it in nodes.txt: user@hostname</p>

<p><b>5</b> find source/dir -type f | parallel --sshloginfile nodes.txt --trc {.}.mp3 lame -a {} -o {.}.mp3 --preset standard --quiet</p>

<p><b>6</b> ./ppss stop -C config.cfg</p>

<p><b>6</b> killall -TERM parallel</p>

<p><b>7</b> ./ppss pause -C config.cfg</p>

<p><b>7</b> Press: CTRL-Z or killall -SIGTSTP parallel</p>

<p><b>8</b> ./ppss continue -C config.cfg</p>

<p><b>8</b> Enter: fg or killall -SIGCONT parallel</p>

<p><b>9</b> ./ppss.sh status -C config.cfg</p>

<p><b>9</b> killall -SIGUSR2 parallel</p>

<h2 id="DIFFERENCES-BETWEEN-pexec-AND-GNU-Parallel">DIFFERENCES BETWEEN pexec AND GNU Parallel</h2>

<p><b>pexec</b> is also a tool for running jobs in parallel.</p>

<h3 id="EXAMPLES-FROM-pexec-MANUAL">EXAMPLES FROM pexec MANUAL</h3>

<p>Here are the examples from <b>pexec</b>&#39;s info page with the equivalent using GNU <b>parallel</b>:</p>

<p><b>1</b> pexec -o sqrt-%s.dat -p &quot;$(seq 10)&quot; -e NUM -n 4 -c -- \ &#39;echo &quot;scale=10000;sqrt($NUM)&quot; | bc&#39;</p>

<p><b>1</b> seq 10 | parallel -j4 &#39;echo &quot;scale=10000;sqrt({})&quot; | bc &gt; sqrt-{}.dat&#39;</p>

<p><b>2</b> pexec -p &quot;$(ls myfiles*.ext)&quot; -i %s -o %s.sort -- sort</p>

<p><b>2</b> ls myfiles*.ext | parallel sort {} &quot;&gt;{}.sort&quot;</p>

<p><b>3</b> pexec -f image.list -n auto -e B -u star.log -c -- \ &#39;fistar $B.fits -f 100 -F id,x,y,flux -o $B.star&#39;</p>

<p><b>3</b> parallel -a image.list \ &#39;fistar {}.fits -f 100 -F id,x,y,flux -o {}.star&#39; 2&gt;star.log</p>

<p><b>4</b> pexec -r *.png -e IMG -c -o - -- \ &#39;convert $IMG ${IMG%.png}.jpeg ; &quot;echo $IMG: done&quot;&#39;</p>

<p><b>4</b> ls *.png | parallel &#39;convert {} {.}.jpeg; echo {}: done&#39;</p>

<p><b>5</b> pexec -r *.png -i %s -o %s.jpg -c &#39;pngtopnm | pnmtojpeg&#39;</p>

<p><b>5</b> ls *.png | parallel &#39;pngtopnm &lt; {} | pnmtojpeg &gt; {}.jpg&#39;</p>

<p><b>6</b> for p in *.png ; do echo ${p%.png} ; done | \ pexec -f - -i %s.png -o %s.jpg -c &#39;pngtopnm | pnmtojpeg&#39;</p>

<p><b>6</b> ls *.png | parallel &#39;pngtopnm &lt; {} | pnmtojpeg &gt; {.}.jpg&#39;</p>

<p><b>7</b> LIST=$(for p in *.png ; do echo ${p%.png} ; done) pexec -r $LIST -i %s.png -o %s.jpg -c &#39;pngtopnm | pnmtojpeg&#39;</p>

<p><b>7</b> ls *.png | parallel &#39;pngtopnm &lt; {} | pnmtojpeg &gt; {.}.jpg&#39;</p>

<p><b>8</b> pexec -n 8 -r *.jpg -y unix -e IMG -c \ &#39;pexec -j -m blockread -d $IMG | \ jpegtopnm | pnmscale 0.5 | pnmtojpeg | \ pexec -j -m blockwrite -s th_$IMG&#39;</p>

<p><b>8</b> Combining GNU <b>parallel</b> and GNU <b>sem</b>.</p>

<p><b>8</b> ls *jpg | parallel -j8 &#39;sem --id blockread cat {} | jpegtopnm |&#39; \ &#39;pnmscale 0.5 | pnmtojpeg | sem --id blockwrite cat &gt; th_{}&#39;</p>

<p><b>8</b> If reading and writing is done to the same disk, this may be faster as only one process will be either reading or writing:</p>

<p><b>8</b> ls *jpg | parallel -j8 &#39;sem --id diskio cat {} | jpegtopnm |&#39; \ &#39;pnmscale 0.5 | pnmtojpeg | sem --id diskio cat &gt; th_{}&#39;</p>

<h2 id="DIFFERENCES-BETWEEN-xjobs-AND-GNU-Parallel">DIFFERENCES BETWEEN xjobs AND GNU Parallel</h2>

<p><b>xjobs</b> is also a tool for running jobs in parallel. It only supports running jobs on your local computer.</p>

<p><b>xjobs</b> deals badly with special characters just like <b>xargs</b>. See the section <b>DIFFERENCES BETWEEN xargs AND GNU Parallel</b>.</p>

<p>Here are the examples from <b>xjobs</b>&#39;s man page with the equivalent using GNU <b>parallel</b>:</p>

<p><b>1</b> ls -1 *.zip | xjobs unzip</p>

<p><b>1</b> ls *.zip | parallel unzip</p>

<p><b>2</b> ls -1 *.zip | xjobs -n unzip</p>

<p><b>2</b> ls *.zip | parallel unzip &gt;/dev/null</p>

<p><b>3</b> find . -name &#39;*.bak&#39; | xjobs gzip</p>

<p><b>3</b> find . -name &#39;*.bak&#39; | parallel gzip</p>

<p><b>4</b> ls -1 *.jar | sed &#39;s/\(.*\)/\1 &gt; \1.idx/&#39; | xjobs jar tf</p>

<p><b>4</b> ls *.jar | parallel jar tf {} &#39;&gt;&#39; {}.idx</p>

<p><b>5</b> xjobs -s script</p>

<p><b>5</b> cat script | parallel</p>

<p><b>6</b> mkfifo /var/run/my_named_pipe; xjobs -s /var/run/my_named_pipe &amp; echo unzip 1.zip &gt;&gt; /var/run/my_named_pipe; echo tar cf /backup/myhome.tar /home/me &gt;&gt; /var/run/my_named_pipe</p>

<p><b>6</b> mkfifo /var/run/my_named_pipe; cat /var/run/my_named_pipe | parallel &amp; echo unzip 1.zip &gt;&gt; /var/run/my_named_pipe; echo tar cf /backup/myhome.tar /home/me &gt;&gt; /var/run/my_named_pipe</p>

<h2 id="DIFFERENCES-BETWEEN-prll-AND-GNU-Parallel">DIFFERENCES BETWEEN prll AND GNU Parallel</h2>

<p><b>prll</b> is also a tool for running jobs in parallel. It does not support running jobs on remote computers.</p>

<p><b>prll</b> encourages using BASH aliases and BASH functions instead of scripts. GNU <b>parallel</b> will never support running aliases (see why http://www.perlmonks.org/index.pl?node_id=484296). However, scripts, composed commands, or functions exported with <b>export -f</b> work just fine.</p>

<p><b>prll</b> generates a lot of status information on stderr (standard error) which makes it harder to use the stderr (standard error) output of the job directly as input for another program.</p>

<p>Here is the example from <b>prll</b>&#39;s man page with the equivalent using GNU <b>parallel</b>:</p>

<p>prll -s &#39;mogrify -flip $1&#39; *.jpg</p>

<p>parallel mogrify -flip ::: *.jpg</p>

<h2 id="DIFFERENCES-BETWEEN-dxargs-AND-GNU-Parallel">DIFFERENCES BETWEEN dxargs AND GNU Parallel</h2>

<p><b>dxargs</b> is also a tool for running jobs in parallel.</p>

<p><b>dxargs</b> does not deal well with more simultaneous jobs than SSHD&#39;s MaxStartups. <b>dxargs</b> is only built for remote run jobs, but does not support transferring of files.</p>

<h2 id="DIFFERENCES-BETWEEN-mdm-middleman-AND-GNU-Parallel">DIFFERENCES BETWEEN mdm/middleman AND GNU Parallel</h2>

<p>middleman(mdm) is also a tool for running jobs in parallel.</p>

<p>Here are the shellscripts of http://mdm.berlios.de/usage.html ported to GNU <b>parallel</b>:</p>

<p><b>seq 19 | parallel buffon -o - | sort -n </b>&gt;<b> result</b></p>

<p><b>cat files | parallel cmd</b></p>

<p><b>find dir -execdir sem cmd {} \;</b></p>

<h2 id="DIFFERENCES-BETWEEN-xapply-AND-GNU-Parallel">DIFFERENCES BETWEEN xapply AND GNU Parallel</h2>

<p><b>xapply</b> can run jobs in parallel on the local computer.</p>

<p>Here are the examples from <b>xapply</b>&#39;s man page with the equivalent using GNU <b>parallel</b>:</p>

<p><b>1</b> xapply &#39;(cd %1 &amp;&amp; make all)&#39; */</p>

<p><b>1</b> parallel &#39;cd {} &amp;&amp; make all&#39; ::: */</p>

<p><b>2</b> xapply -f &#39;diff %1 ../version5/%1&#39; manifest | more</p>

<p><b>2</b> parallel diff {} ../version5/{} &lt; manifest | more</p>

<p><b>3</b> xapply -p/dev/null -f &#39;diff %1 %2&#39; manifest1 checklist1</p>

<p><b>3</b> parallel --xapply diff {1} {2} :::: manifest1 checklist1</p>

<p><b>4</b> xapply &#39;indent&#39; *.c</p>

<p><b>4</b> parallel indent ::: *.c</p>

<p><b>5</b> find ~ksb/bin -type f ! -perm -111 -print | xapply -f -v &#39;chmod a+x&#39; -</p>

<p><b>5</b> find ~ksb/bin -type f ! -perm -111 -print | parallel -v chmod a+x</p>

<p><b>6</b> find */ -... | fmt 960 1024 | xapply -f -i /dev/tty &#39;vi&#39; -</p>

<p><b>6</b> sh &lt;(find */ -... | parallel -s 1024 echo vi)</p>

<p><b>6</b> find */ -... | parallel -s 1024 -Xuj1 vi</p>

<p><b>7</b> find ... | xapply -f -5 -i /dev/tty &#39;vi&#39; - - - - -</p>

<p><b>7</b> sh &lt;(find ... |parallel -n5 echo vi)</p>

<p><b>7</b> find ... |parallel -n5 -uj1 vi</p>

<p><b>8</b> xapply -fn &quot;&quot; /etc/passwd</p>

<p><b>8</b> parallel -k echo &lt; /etc/passwd</p>

<p><b>9</b> tr &#39;:&#39; &#39;\012&#39; &lt; /etc/passwd | xapply -7 -nf &#39;chown %1 %6&#39; - - - - - - -</p>

<p><b>9</b> tr &#39;:&#39; &#39;\012&#39; &lt; /etc/passwd | parallel -N7 chown {1} {6}</p>

<p><b>10</b> xapply &#39;[ -d %1/RCS ] || echo %1&#39; */</p>

<p><b>10</b> parallel &#39;[ -d {}/RCS ] || echo {}&#39; ::: */</p>

<p><b>11</b> xapply -f &#39;[ -f %1 ] &amp;&amp; echo %1&#39; List | ...</p>

<p><b>11</b> parallel &#39;[ -f {} ] &amp;&amp; echo {}&#39; &lt; List | ...</p>

<h2 id="DIFFERENCES-BETWEEN-paexec-AND-GNU-Parallel">DIFFERENCES BETWEEN paexec AND GNU Parallel</h2>

<p><b>paexec</b> can run jobs in parallel on both the local and remote computers.</p>

<p><b>paexec</b> requires commands to print a blank line as the last output. This means you will have to write a wrapper for most programs.</p>

<p><b>paexec</b> has a job dependency facility so a job can depend on another job to be executed successfully. Sort of a poor-man&#39;s <b>make</b>.</p>

<p>Here are the examples from <b>paexec</b>&#39;s example catalog with the equivalent using GNU <b>parallel</b>:</p>

<dl>

<dt id="div_X_run">1_div_X_run:</dt>
<dd>

<pre><code>  ../../paexec -s -l -c &quot;`pwd`/1_div_X_cmd&quot; -n +1 &lt;&lt;EOF [...]
  parallel echo {} &#39;|&#39; `pwd`/1_div_X_cmd &lt;&lt;EOF [...]</code></pre>

</dd>
<dt id="all_substr_run">all_substr_run:</dt>
<dd>

<pre><code>  ../../paexec -lp -c &quot;`pwd`/all_substr_cmd&quot; -n +3 &lt;&lt;EOF [...]
  parallel echo {} &#39;|&#39; `pwd`/all_substr_cmd &lt;&lt;EOF [...]</code></pre>

</dd>
<dt id="cc_wrapper_run">cc_wrapper_run:</dt>
<dd>

<pre><code>  ../../paexec -c &quot;env CC=gcc CFLAGS=-O2 `pwd`/cc_wrapper_cmd&quot; \
             -n &#39;host1 host2&#39; \
             -t &#39;/usr/bin/ssh -x&#39; &lt;&lt;EOF [...]
  parallel echo {} &#39;|&#39; &quot;env CC=gcc CFLAGS=-O2 `pwd`/cc_wrapper_cmd&quot; \
             -S host1,host2 &lt;&lt;EOF [...]
  # This is not exactly the same, but avoids the wrapper
  parallel gcc -O2 -c -o {.}.o {} \
             -S host1,host2 &lt;&lt;EOF [...]</code></pre>

</dd>
<dt id="toupper_run">toupper_run:</dt>
<dd>

<pre><code>  ../../paexec -lp -c &quot;`pwd`/toupper_cmd&quot; -n +10 &lt;&lt;EOF [...]
  parallel echo {} &#39;|&#39; ./toupper_cmd &lt;&lt;EOF [...]
  # Without the wrapper:
  parallel echo {} &#39;| awk {print\ toupper\(\$0\)}&#39; &lt;&lt;EOF [...]</code></pre>

</dd>
</dl>

<h2 id="DIFFERENCES-BETWEEN-map-AND-GNU-Parallel">DIFFERENCES BETWEEN map AND GNU Parallel</h2>

<p><b>map</b> sees it as a feature to have less features and in doing so it also handles corner cases incorrectly. A lot of GNU <b>parallel</b>&#39;s code is to handle corner cases correctly on every platform, so you will not get a nasty surprise if a user for example saves a file called: <i>My brother&#39;s 12&quot; records.txt</i></p>

<p><b>map</b>&#39;s example showing how to deal with special characters fails on special characters:</p>

<pre><code>  echo &quot;The Cure&quot; &gt; My\ brother\&#39;s\ 12\&quot;\ records

  ls | map &#39;echo -n `gzip &lt; &quot;%&quot; | wc -c`; echo -n &#39;*100/&#39;; wc -c &lt; &quot;%&quot;&#39; | bc</code></pre>

<p>It works with GNU <b>parallel</b>:</p>

<pre><code>  ls | parallel &#39;echo -n `gzip &lt; {} | wc -c`; echo -n &#39;*100/&#39;; wc -c &lt; {}&#39; | bc</code></pre>

<p>And you can even get the file name prepended:</p>

<pre><code>  ls | parallel --tag &#39;(echo -n `gzip &lt; {} | wc -c`&#39;*100/&#39;; wc -c &lt; {}) | bc&#39;</code></pre>

<p><b>map</b> has no support for grouping. So this gives the wrong results without any warnings:</p>

<pre><code>  parallel perl -e &#39;\$a=\&quot;1{}\&quot;x10000000\;print\ \$a,\&quot;\\n\&quot;&#39; &#39;&gt;&#39; {} ::: a b c d e f
  ls -l a b c d e f
  parallel -kP4 -n1 grep 1 &gt; out.par ::: a b c d e f
  map -p 4 &#39;grep 1&#39; a b c d e f &gt; out.map-unbuf
  map -p 4 &#39;grep --line-buffered 1&#39; a b c d e f &gt; out.map-linebuf
  map -p 1 &#39;grep --line-buffered 1&#39; a b c d e f &gt; out.map-serial
  ls -l out*
  md5sum out*</code></pre>

<p>The documentation shows a workaround, but not only does that mix stdout (standard output) with stderr (standard error) it also fails completely for certain jobs (and may even be considered less readable):</p>

<pre><code>  parallel echo -n {} ::: 1 2 3

  map -p 4 &#39;echo -n % 2&gt;&amp;1 | sed -e &quot;s/^/$$:/&quot;&#39; 1 2 3 | sort | cut -f2- -d:</code></pre>

<p><b>map</b> cannot handle bundled options: <b>map -vp 0 echo this fails</b></p>

<p><b>map</b> does not have an argument separator on the command line, but uses the first argument as command. This makes quoting harder which again may affect readability. Compare:</p>

<pre><code>  map -p 2 perl\\\ -ne\\\ \\\&#39;/^\\\\S+\\\\s+\\\\S+\\\$/\\\ and\\\ print\\\ \\\$ARGV,\\\&quot;\\\\n\\\&quot;\\\&#39; *

  parallel -q perl -ne &#39;/^\S+\s+\S+$/ and print $ARGV,&quot;\n&quot;&#39; ::: *</code></pre>

<p><b>map</b> can do multiple arguments with context replace, but not without context replace:</p>

<pre><code>  parallel --xargs echo &#39;BEGIN{&#39;{}&#39;}END&#39; ::: 1 2 3</code></pre>

<p><b>map</b> does not set exit value according to whether one of the jobs failed:</p>

<pre><code>  parallel false ::: 1 || echo Job failed

  map false 1 || echo Never run</code></pre>

<p><b>map</b> requires Perl v5.10.0 making it harder to use on old systems.</p>

<p><b>map</b> has no way of using % in the command (GNU Parallel has -I to specify another replacement string than {}).</p>

<p>By design <b>map</b> is option incompatible with <b>xargs</b>, it does not have remote job execution, a structured way of saving results, multiple input sources, progress indicator, configurable record delimiter (only field delimiter), logging of jobs run with possibility to resume, keeping the output in the same order as input, --pipe processing, and dynamically timeouts.</p>

<h2 id="DIFFERENCES-BETWEEN-ladon-AND-GNU-Parallel">DIFFERENCES BETWEEN ladon AND GNU Parallel</h2>

<p><b>ladon</b> can run multiple jobs on files in parallel.</p>

<p><b>ladon</b> only works on files and the only way to specify files is using a quoted glob string (such as \*.jpg). It is not possible to list the files manually.</p>

<p>As replacement strings it uses FULLPATH DIRNAME BASENAME EXT RELDIR RELPATH</p>

<p>These can be simulated using GNU <b>parallel</b> by putting this in <b>~/.parallel/config</b>:</p>

<pre><code>    --rpl &#39;FULLPATH $_=::shell_quote($_);chomp($_=qx{readlink -f $_});&#39;
    --rpl &#39;DIRNAME $_=::shell_quote(::dirname($_));chomp($_=qx{readlink -f $_});&#39;
    --rpl &#39;BASENAME s:.*/::;s:\.[^/.]+$::;&#39;
    --rpl &#39;EXT s:.*\.::&#39;
    --rpl &#39;RELDIR $_=::shell_quote($_);chomp(($_,$c)=qx{readlink -f $_;pwd});s:\Q$c/\E::;$_=::dirname($_);&#39;
    --rpl &#39;RELPATH $_=::shell_quote($_);chomp(($_,$c)=qx{readlink -f $_;pwd});s:\Q$c/\E::;&#39;</code></pre>

<p><b>ladon</b> deals badly with filenames containing &quot; and newline, and it fails for output larger than 200k:</p>

<pre><code>    ladon &#39;*&#39; -- seq 36000 | wc</code></pre>

<h3 id="EXAMPLES-FROM-ladon-MANUAL">EXAMPLES FROM ladon MANUAL</h3>

<p>It is assumed that the &#39;--rpl&#39;s above are put in <b>~/.parallel/config</b> and that it is run under a shell that supports &#39;**&#39; globbing (such as <b>zsh</b>):</p>

<p><b>1</b> ladon &quot;**/*.txt&quot; -- echo RELPATH</p>

<p><b>1</b> parallel echo RELPATH ::: **/*.txt</p>

<p><b>2</b> ladon &quot;~/Documents/**/*.pdf&quot; -- shasum FULLPATH &gt;hashes.txt</p>

<p><b>2</b> parallel shasum FULLPATH ::: ~/Documents/**/*.pdf &gt;hashes.txt</p>

<p><b>3</b> ladon -m thumbs/RELDIR &quot;**/*.jpg&quot; -- convert FULLPATH -thumbnail 100x100^ -gravity center -extent 100x100 thumbs/RELPATH</p>

<p><b>3</b> parallel mkdir -p thumbs/RELDIR\; convert FULLPATH -thumbnail 100x100^ -gravity center -extent 100x100 thumbs/RELPATH ::: **/*.jpg</p>

<p><b>4</b> ladon &quot;~/Music/*.wav&quot; -- lame -V 2 FULLPATH DIRNAME/BASENAME.mp3</p>

<p><b>4</b> parallel lame -V 2 FULLPATH DIRNAME/BASENAME.mp3 ::: ~/Music/*.wav</p>

<h2 id="DIFFERENCES-BETWEEN-ClusterSSH-AND-GNU-Parallel">DIFFERENCES BETWEEN ClusterSSH AND GNU Parallel</h2>

<p>ClusterSSH solves a different problem than GNU <b>parallel</b>.</p>

<p>ClusterSSH opens a terminal window for each computer and using a master window you can run the same command on all the computers. This is typically used for administrating several computers that are almost identical.</p>

<p>GNU <b>parallel</b> runs the same (or different) commands with different arguments in parallel possibly using remote computers to help computing. If more than one computer is listed in <b>-S</b> GNU <b>parallel</b> may only use one of these (e.g. if there are 8 jobs to be run and one computer has 8 cores).</p>

<p>GNU <b>parallel</b> can be used as a poor-man&#39;s version of ClusterSSH:</p>

<p><b>parallel --nonall -S server-a,server-b do_stuff foo bar</b></p>

<h1 id="BUGS">BUGS</h1>

<h2 id="Quoting-of-newline">Quoting of newline</h2>

<p>Because of the way newline is quoted this will not work:</p>

<p>echo 1,2,3 | parallel -vkd, &quot;echo &#39;a{}b&#39;&quot;</p>

<p>However, these will all work:</p>

<p>echo 1,2,3 | parallel -vkd, echo a{}b</p>

<p>echo 1,2,3 | parallel -vkd, &quot;echo &#39;a&#39;{}&#39;b&#39;&quot;</p>

<p>echo 1,2,3 | parallel -vkd, &quot;echo &#39;a&#39;&quot;{}&quot;&#39;b&#39;&quot;</p>

<h2 id="Speed">Speed</h2>

<h3 id="Startup">Startup</h3>

<p>GNU <b>parallel</b> is slow at starting up - around 250 ms the first time and 150 ms after that.</p>

<h3 id="Job-startup">Job startup</h3>

<p>Starting a job on the local machine takes around 10 ms. This can be a big overhead if the job takes very few ms to run. Often you can group small jobs together using <b>-X</b> which will make the overhead less significant. Or you can run multiple GNU <b>parallel</b>s as described in <b>EXAMPLE: Speeding up fast jobs</b>.</p>

<h3 id="SSH">SSH</h3>

<p>When using multiple computers GNU <b>parallel</b> opens <b>ssh</b> connections to them to figure out how many connections can be used reliably simultaneously (Namely SSHD&#39;s MaxStartups). This test is done for each host in serial, so if your <b>--sshloginfile</b> contains many hosts it may be slow.</p>

<p>If your jobs are short you may see that there are fewer jobs running on the remove systems than expected. This is due to time spent logging in and out. <b>-M</b> may help here.</p>

<h3 id="Disk-access">Disk access</h3>

<p>A single disk can normally read data faster if it reads one file at a time instead of reading a lot of files in parallel, as this will avoid disk seeks. However, newer disk systems with multiple drives can read faster if reading from multiple files in parallel.</p>

<p>If the jobs are of the form read-all-compute-all-write-all, so everything is read before anything is written, it may be faster to force only one disk access at the time:</p>

<pre><code>  sem --id diskio cat file | compute | sem --id diskio cat &gt; file</code></pre>

<p>If the jobs are of the form read-compute-write, so writing starts before all reading is done, it may be faster to force only one reader and writer at the time:</p>

<pre><code>  sem --id read cat file | compute | sem --id write cat &gt; file</code></pre>

<p>If the jobs are of the form read-compute-read-compute, it may be faster to run more jobs in parallel than the system has CPUs, as some of the jobs will be stuck waiting for disk access.</p>

<h2 id="nice-limits-command-length">--nice limits command length</h2>

<p>The current implementation of <b>--nice</b> is too pessimistic in the max allowed command length. It only uses a little more than half of what it could. This affects <b>-X</b> and <b>-m</b>. If this becomes a real problem for you file a bug-report.</p>

<h2 id="Aliases-and-functions-do-not-work">Aliases and functions do not work</h2>

<p>If you get:</p>

<p><b>Can&#39;t exec &quot;<i>command</i>&quot;: No such file or directory</b></p>

<p>or:</p>

<p><b>open3: exec of by <i>command</i> failed</b></p>

<p>it may be because <i>command</i> is not known, but it could also be because <i>command</i> is an alias or a function. If it is a function you need to <b>export -f</b> the function first. An alias will, however, not work (see why http://www.perlmonks.org/index.pl?node_id=484296), so change your alias to a script.</p>

<h1 id="REPORTING-BUGS">REPORTING BUGS</h1>

<p>Report bugs to &lt;bug-parallel@gnu.org&gt; or https://savannah.gnu.org/bugs/?func=additem&amp;group=parallel</p>

<p>Your bug report should always include:</p>

<ul>

<li><p>The error message you get (if any).</p>

</li>
<li><p>The complete output of <b>parallel --version</b>. If you are not running the latest released version you should specify why you believe the problem is not fixed in that version.</p>

</li>
<li><p>A complete example that others can run that shows the problem. This should preferably be small and simple. A combination of <b>yes</b>, <b>seq</b>, <b>cat</b>, <b>echo</b>, and <b>sleep</b> can reproduce most errors. If your example requires large files, see if you can make them by something like <b>seq 1000000</b> &gt; <b>file</b> or <b>yes | head -n 10000000</b> &gt; <b>file</b>. If your example requires remote execution, see if you can use <b>localhost</b> - maybe using another login.</p>

</li>
<li><p>The output of your example. If your problem is not easily reproduced by others, the output might help them figure out the problem.</p>

</li>
<li><p>Whether you have watched the intro videos (http://www.youtube.com/playlist?list=PL284C9FF2488BC6D1), walked through the tutorial (man parallel_tutorial), and read the EXAMPLE section in the man page (man parallel - search for EXAMPLE:).</p>

</li>
</ul>

<p>If you suspect the error is dependent on your environment or distribution, please see if you can reproduce the error on one of these VirtualBox images: http://sourceforge.net/projects/virtualboximage/files/</p>

<p>Specifying the name of your distribution is not enough as you may have installed software that is not in the VirtualBox images.</p>

<p>If you cannot reproduce the error on any of the VirtualBox images above, you should assume the debugging will be done through you. That will put more burden on you and it is extra important you give any information that help. In general the problem will be fixed faster and with less work for you if you can reproduce the error on a VirtualBox.</p>

<h1 id="AUTHOR">AUTHOR</h1>

<p>When using GNU <b>parallel</b> for a publication please cite:</p>

<p>O. Tange (2011): GNU Parallel - The Command-Line Power Tool, ;login: The USENIX Magazine, February 2011:42-47.</p>

<p>Alternatively you can get GNU Parallel without this requirement by paying 10000 EUR.</p>

<p>Copyright (C) 2007-10-18 Ole Tange, http://ole.tange.dk</p>

<p>Copyright (C) 2008,2009,2010 Ole Tange, http://ole.tange.dk</p>

<p>Copyright (C) 2010,2011,2012,2013,2014 Ole Tange, http://ole.tange.dk and Free Software Foundation, Inc.</p>

<p>Parts of the manual concerning <b>xargs</b> compatibility is inspired by the manual of <b>xargs</b> from GNU findutils 4.4.2.</p>

<h1 id="LICENSE">LICENSE</h1>

<p>Copyright (C) 2007,2008,2009,2010,2011,2012,2013 Free Software Foundation, Inc.</p>

<p>This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or at your option any later version.</p>

<p>This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.</p>

<p>You should have received a copy of the GNU General Public License along with this program. If not, see &lt;http://www.gnu.org/licenses/&gt;.</p>

<h2 id="Documentation-license-I">Documentation license I</h2>

<p>Permission is granted to copy, distribute and/or modify this documentation under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, with no Front-Cover Texts, and with no Back-Cover Texts. A copy of the license is included in the file fdl.txt.</p>

<h2 id="Documentation-license-II">Documentation license II</h2>

<p>You are free:</p>

<dl>

<dt id="to-Share"><b>to Share</b></dt>
<dd>

<p>to copy, distribute and transmit the work</p>

</dd>
<dt id="to-Remix"><b>to Remix</b></dt>
<dd>

<p>to adapt the work</p>

</dd>
</dl>

<p>Under the following conditions:</p>

<dl>

<dt id="Attribution"><b>Attribution</b></dt>
<dd>

<p>You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work).</p>

</dd>
<dt id="Share-Alike"><b>Share Alike</b></dt>
<dd>

<p>If you alter, transform, or build upon this work, you may distribute the resulting work only under the same, similar or a compatible license.</p>

</dd>
</dl>

<p>With the understanding that:</p>

<dl>

<dt id="Waiver"><b>Waiver</b></dt>
<dd>

<p>Any of the above conditions can be waived if you get permission from the copyright holder.</p>

</dd>
<dt id="Public-Domain"><b>Public Domain</b></dt>
<dd>

<p>Where the work or any of its elements is in the public domain under applicable law, that status is in no way affected by the license.</p>

</dd>
<dt id="Other-Rights"><b>Other Rights</b></dt>
<dd>

<p>In no way are any of the following rights affected by the license:</p>

<ul>

<li><p>Your fair dealing or fair use rights, or other applicable copyright exceptions and limitations;</p>

</li>
<li><p>The author&#39;s moral rights;</p>

</li>
<li><p>Rights other persons may have either in the work itself or in how the work is used, such as publicity or privacy rights.</p>

</li>
</ul>

</dd>
</dl>

<dl>

<dt id="Notice"><b>Notice</b></dt>
<dd>

<p>For any reuse or distribution, you must make clear to others the license terms of this work.</p>

</dd>
</dl>

<p>A copy of the full license is included in the file as cc-by-sa.txt.</p>

<h1 id="DEPENDENCIES">DEPENDENCIES</h1>

<p>GNU <b>parallel</b> uses Perl, and the Perl modules Getopt::Long, IPC::Open3, Symbol, IO::File, POSIX, and File::Temp. For remote usage it also uses rsync with ssh.</p>

<h1 id="SEE-ALSO">SEE ALSO</h1>

<p><b>ssh</b>(1), <b>rsync</b>(1), <b>find</b>(1), <b>xargs</b>(1), <b>dirname</b>(1), <b>make</b>(1), <b>pexec</b>(1), <b>ppss</b>(1), <b>xjobs</b>(1), <b>prll</b>(1), <b>dxargs</b>(1), <b>mdm</b>(1)</p>


</body>

</html>