/usr/share/doc/condor/html/3_11_High_Availability.html is in condor-doc 7.8.2~dfsg.1-1+deb7u3.
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 | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<!--Converted with LaTeX2HTML 2008 (1.71)
original version by: Nikos Drakos, CBLU, University of Leeds
* revised and updated by: Marcus Hennecke, Ross Moore, Herb Swan
* with significant contributions from:
Jens Lippmann, Marek Rouchal, Martin Wilck and others -->
<HTML>
<HEAD>
<TITLE>3.11 The High Availability of Daemons</TITLE>
<META NAME="description" CONTENT="3.11 The High Availability of Daemons">
<META NAME="keywords" CONTENT="ref">
<META NAME="resource-type" CONTENT="document">
<META NAME="distribution" CONTENT="global">
<META NAME="Generator" CONTENT="LaTeX2HTML v2008">
<META HTTP-EQUIV="Content-Style-Type" CONTENT="text/css">
<LINK REL="STYLESHEET" HREF="ref.css">
<LINK REL="next" HREF="3_12_Setting_Up.html">
<LINK REL="previous" HREF="3_10_Pool_Management.html">
<LINK REL="up" HREF="3_Administrators_Manual.html">
<LINK REL="next" HREF="3_12_Setting_Up.html">
</HEAD>
<BODY BGCOLOR=#FFFFFF >
<DIV CLASS="navigation"><!--Navigation Panel-->
<A NAME="tex2html1587"
HREF="3_12_Setting_Up.html">
<IMG WIDTH="37" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="next" SRC="next.png"></A>
<A NAME="tex2html1581"
HREF="3_Administrators_Manual.html">
<IMG WIDTH="26" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="up" SRC="up.png"></A>
<A NAME="tex2html1575"
HREF="3_10_Pool_Management.html">
<IMG WIDTH="63" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="previous" SRC="prev.png"></A>
<A NAME="tex2html1583"
HREF="Contents.html">
<IMG WIDTH="65" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="contents" SRC="contents.png"></A>
<A NAME="tex2html1585"
HREF="Index.html">
<IMG WIDTH="43" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="index" SRC="index.png"></A>
<BR>
<B> Next:</B> <A NAME="tex2html1588"
HREF="3_12_Setting_Up.html">3.12 Setting Up for</A>
<B> Up:</B> <A NAME="tex2html1582"
HREF="3_Administrators_Manual.html">3. Administrators' Manual</A>
<B> Previous:</B> <A NAME="tex2html1576"
HREF="3_10_Pool_Management.html">3.10 Pool Management</A>
<B> <A NAME="tex2html1584"
HREF="Contents.html">Contents</A></B>
<B> <A NAME="tex2html1586"
HREF="Index.html">Index</A></B>
<BR>
<BR></DIV>
<!--End of Navigation Panel-->
<!--Table of Child-Links-->
<A NAME="CHILD_LINKS"><STRONG>Subsections</STRONG></A>
<UL CLASS="ChildLinks">
<LI><A NAME="tex2html1589"
HREF="3_11_High_Availability.html#SECTION004111000000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">1</SPAN> High Availability of the Job Queue</A>
<UL>
<LI><A NAME="tex2html1590"
HREF="3_11_High_Availability.html#SECTION004111100000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">1</SPAN>.<SPAN CLASS="arabic">1</SPAN> Working with Remote Job Submission</A>
</UL>
<BR>
<LI><A NAME="tex2html1591"
HREF="3_11_High_Availability.html#SECTION004112000000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN> High Availability of the
Central Manager</A>
<UL>
<LI><A NAME="tex2html1592"
HREF="3_11_High_Availability.html#SECTION004112100000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">1</SPAN> Interaction with Flocking</A>
<LI><A NAME="tex2html1593"
HREF="3_11_High_Availability.html#SECTION004112200000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">2</SPAN> Introduction</A>
<LI><A NAME="tex2html1594"
HREF="3_11_High_Availability.html#SECTION004112300000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">3</SPAN> Configuration</A>
<LI><A NAME="tex2html1595"
HREF="3_11_High_Availability.html#SECTION004112400000000000000"><SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">4</SPAN> Sample Configuration</A>
</UL></UL>
<!--End of Table of Child-Links-->
<HR>
<H1><A NAME="SECTION004110000000000000000"></A><A NAME="sec:high-availability"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN> The High Availability of Daemons
</H1>
<P>
<A NAME="34192"></A>
<P>
In the case that a key machine no longer functions,
Condor can be configured such that another machine takes on
the key functions.
This is called <SPAN CLASS="textit">High Availability</SPAN>.
While high availability is generally applicable,
there are currently two specialized cases for its use:
when the central manager
(running the <SPAN CLASS="textit">condor_negotiator</SPAN> and <SPAN CLASS="textit">condor_collector</SPAN> daemons)
becomes unavailable,
and when the machine running the <SPAN CLASS="textit">condor_schedd</SPAN>
daemon (maintaining the job queue) becomes unavailable.
<P>
<H2><A NAME="SECTION004111000000000000000"></A><A NAME="sec:HA-schedd"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">1</SPAN> High Availability of the Job Queue
</H2>
<P>
<A NAME="34211"></A>
<P>
For a pool where all jobs are submitted through
a single machine in the pool,
and there are lots of jobs,
this machine becoming nonfunctional means that
jobs stop running.
The <SPAN CLASS="textit">condor_schedd</SPAN> daemon maintains the job queue.
No job queue due to having a nonfunctional machine
implies that no jobs can be run.
This situation is worsened by using one machine
as the single submission point.
For each Condor job (taken from the queue) that is executed,
a <SPAN CLASS="textit">condor_shadow</SPAN> process runs on the machine where submitted
to handle input/output functionality.
If this machine becomes nonfunctional, none of the jobs can
continue.
The entire pool stops running jobs.
<P>
The goal of <SPAN CLASS="textit">High Availability</SPAN> in this special case is
to transfer the <SPAN CLASS="textit">condor_schedd</SPAN> daemon to run on another
designated machine.
Jobs caused to stop without finishing can be restarted from the
beginning, or can continue execution using the most recent checkpoint.
New jobs can enter the job queue.
Without <SPAN CLASS="textit">High Availability</SPAN>,
the job queue would remain intact, but further progress on jobs
would wait until the machine running the <SPAN CLASS="textit">condor_schedd</SPAN> daemon
became available (after fixing whatever caused it to become
unavailable).
<P>
Condor uses its flexible configuration mechanisms to allow
the transfer of the <SPAN CLASS="textit">condor_schedd</SPAN> daemon from one machine
to another.
The configuration specifies
which machines are chosen to run the <SPAN CLASS="textit">condor_schedd</SPAN> daemon.
To prevent multiple <SPAN CLASS="textit">condor_schedd</SPAN> daemons from running at the same time,
a lock (semaphore-like) is held over the job queue.
This synchronizes the situation in which control is
transferred to a secondary machine,
and the primary machine returns to functionality.
Configuration variables also determine time intervals at which
the lock expires,
and periods of time that pass between polling to check
for expired locks.
<P>
To specify a single machine that would take over, if the
machine running the <SPAN CLASS="textit">condor_schedd</SPAN> daemon stops working,
the following additions are made to the local configuration
of any and all machines that are able to run the <SPAN CLASS="textit">condor_schedd</SPAN> daemon
(becoming the single pool submission point):
<P>
<PRE>
MASTER_HA_LIST = SCHEDD
SPOOL = /share/spool
HA_LOCK_URL = file:/share/spool
VALID_SPOOL_FILES = $(VALID_SPOOL_FILES), SCHEDD.lock
</PRE>
<P>
Configuration macro <TT>MASTER_HA_LIST</TT> <A NAME="34393"></A> <A NAME="34394"></A> identifies the
<SPAN CLASS="textit">condor_schedd</SPAN> daemon as the daemon that is to be watched
to make sure that it is running.
Each machine with this configuration must have access to the
lock (the job queue) which synchronizes which single machine does run the
<SPAN CLASS="textit">condor_schedd</SPAN> daemon.
This lock and the job queue must both be located in a shared file space,
and is currently specified only with a file URL.
The configuration specifies the shared space
(<TT>SPOOL</TT>),
and the URL of the lock.
<SPAN CLASS="textit">condor_preen</SPAN> is not currently aware of the lock file and will
delete it if it is placed in the <TT>SPOOL</TT> directory, so be
sure to add SCHEDD.lock to <TT>VALID_SPOOL_FILES</TT> <A NAME="34406"></A> <A NAME="34407"></A>.
<P>
As Condor starts on machines that are configured to run
the single <SPAN CLASS="textit">condor_schedd</SPAN> daemon,
the <SPAN CLASS="textit">condor_master</SPAN> daemon of the
first machine that looks at (polls) the lock
and notices that no lock is held.
This implies that no <SPAN CLASS="textit">condor_schedd</SPAN> daemon is running.
This <SPAN CLASS="textit">condor_master</SPAN> daemon acquires the lock
and runs the <SPAN CLASS="textit">condor_schedd</SPAN> daemon.
Other machines with this same capability to run the
<SPAN CLASS="textit">condor_schedd</SPAN> daemon look at (poll) the lock,
but do not run the daemon, as the lock is held.
The machine running the <SPAN CLASS="textit">condor_schedd</SPAN> daemon renews the
lock periodically.
<P>
If the machine running the <SPAN CLASS="textit">condor_schedd</SPAN> daemon fails to renew
the lock (because the machine is not functioning),
the lock times out (becomes stale).
The lock is released by the <SPAN CLASS="textit">condor_master</SPAN> daemon
if <SPAN CLASS="textit">condor_off</SPAN> or <SPAN CLASS="textit">condor_off -schedd</SPAN> is
executed, or when the <SPAN CLASS="textit">condor_master</SPAN> daemon knows that the
<SPAN CLASS="textit">condor_schedd</SPAN> daemon is no longer running.
As other machines capable of running the <SPAN CLASS="textit">condor_schedd</SPAN> daemon
look at the lock (poll), one machine will be the first
to notice that the lock has timed out or been released.
This machine (correctly) interprets this situation as the
<SPAN CLASS="textit">condor_schedd</SPAN> daemon is no longer running.
This machine's <SPAN CLASS="textit">condor_master</SPAN> daemon then acquires the lock
and runs the <SPAN CLASS="textit">condor_schedd</SPAN> daemon.
<P>
See
section <A HREF="3_3_Configuration.html#sec:Master-Config-File-Entries">3.3.9</A>,
in the section on <SPAN CLASS="textit">condor_master</SPAN> Configuration File Macros
for details relating to the configuration variables
used to set timing and polling intervals.
<P>
<H3><A NAME="SECTION004111100000000000000"></A><A NAME="sec:HA-remote-submit"></A>
<A NAME="34252"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">1</SPAN>.<SPAN CLASS="arabic">1</SPAN> Working with Remote Job Submission
</H3>
<P>
Remote job submission requires identification of the job queue,
submitting with a command similar to:
<PRE>
% condor_submit -remote condor@example.com myjob.submit
</PRE>This implies the identification of a single <SPAN CLASS="textit">condor_schedd</SPAN> daemon,
running on a single machine.
With the high availability of the job queue, there are
multiple <SPAN CLASS="textit">condor_schedd</SPAN> daemons, of which only one at a time is acting
as the single submission point.
To make remote submission of jobs work properly,
set the configuration variable <TT>SCHEDD_NAME</TT> <A NAME="34450"></A> <A NAME="34451"></A> in the local
configuration to have
the same value for each potentially running <SPAN CLASS="textit">condor_schedd</SPAN> daemon.
In addition, the value chosen for the variable <TT>SCHEDD_NAME</TT>
will need to include the at symbol (<code>@</code>),
such that Condor will not modify the value set for this variable.
See the description of <TT>MASTER_NAME</TT> in
section <A HREF="3_3_Configuration.html#param:MasterName">3.3.9</A> on page <A HREF="3_3_Configuration.html#param:MasterName"><IMG ALIGN="BOTTOM" BORDER="1" ALT="[*]" SRC="crossref.png"></A>
for defaults and composition of valid values for <TT>SCHEDD_NAME</TT>.
As an example, include in each local configuration a value similar to:
<PRE>
SCHEDD_NAME = had-schedd@
</PRE>
<P>
Then, with this sample configuration, the
submit command appears as:
<PRE>
% condor_submit -remote had-schedd@ myjob.submit
</PRE>
<P>
<H2><A NAME="SECTION004112000000000000000"></A><A NAME="sec:HA-negotiator"></A>
<A NAME="34269"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN> High Availability of the
Central Manager
</H2>
<P>
<H3><A NAME="SECTION004112100000000000000"></A><A NAME="sec:HA-not-with-flocking"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">1</SPAN> Interaction with Flocking
</H3>
<P>
The Condor high availability mechanisms discussed in this section
currently do not work well in configurations involving flocking.
The individual problems listed listed below interact to make the situation
worse.
Because of these problems, we advise against the use of
flocking to pools with high availability mechanisms enabled.
<P>
<UL>
<LI>The <SPAN CLASS="textit">condor_schedd</SPAN> has a hard configured list of
<SPAN CLASS="textit">condor_collector</SPAN> and <SPAN CLASS="textit">condor_negotiator</SPAN> daemons,
and does not query
redundant collectors to get the current <SPAN CLASS="textit">condor_negotiator</SPAN>,
as it does when communicating with its local pool.
As a result, if the
default <SPAN CLASS="textit">condor_negotiator</SPAN> fails, the <SPAN CLASS="textit">condor_schedd</SPAN> does not learn
of the failure,
and thus, talk to the new <SPAN CLASS="textit">condor_negotiator</SPAN>.
<P>
</LI>
<LI>When the <SPAN CLASS="textit">condor_negotiator</SPAN> is unable to communicate with a
<SPAN CLASS="textit">condor_collector</SPAN>, it utilizes the next <SPAN CLASS="textit">condor_collector</SPAN>
within the list.
Unfortunately, it does not start over at the top of the list.
When combined with the previous problem,
a backup <SPAN CLASS="textit">condor_negotiator</SPAN> will never get
jobs from a flocked <SPAN CLASS="textit">condor_schedd</SPAN>.
<P>
</LI>
</UL>
<P>
<H3><A NAME="SECTION004112200000000000000"></A><A NAME="sec:HA-of-CM-intro"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">2</SPAN> Introduction
</H3>
<P>
The <SPAN CLASS="textit">condor_negotiator</SPAN> and <SPAN CLASS="textit">condor_collector</SPAN> daemons
are the heart of the Condor matchmaking system.
The availability of these daemons is critical to a Condor pool's
functionality.
Both daemons usually run on the same machine,
most often known as the central manager.
The failure of a central manager machine prevents Condor from matching
new jobs and allocating new resources.
High availability of the <SPAN CLASS="textit">condor_negotiator</SPAN> and <SPAN CLASS="textit">condor_collector</SPAN> daemons
eliminates this problem.
<P>
Configuration allows one of
multiple machines within the pool to function as the central manager.
While there are may be many active <SPAN CLASS="textit">condor_collector</SPAN> daemons,
only a single, active <SPAN CLASS="textit">condor_negotiator</SPAN> daemon will be running.
The machine with the <SPAN CLASS="textit">condor_negotiator</SPAN> daemon running is
the active central manager.
The other potential central managers each
have a <SPAN CLASS="textit">condor_collector</SPAN> daemon running;
these are the idle central managers.
<P>
All submit and execute machines are configured to report
to all potential central manager machines.
<P>
<A NAME="34366"></A>
<A NAME="34367"></A>
<A NAME="34296"></A>
Each potential central manager machine runs the high
availability daemon, <SPAN CLASS="textit">condor_had</SPAN>.
These daemons communicate with each other,
constantly monitoring the pool to ensure that one active central
manager is available.
If the active central manager machine crashes or is shut down,
these daemons detect the failure,
and they agree on
which of the idle central managers is to become the active one.
A protocol determines this.
<P>
In the case of a network partition,
idle <SPAN CLASS="textit">condor_had</SPAN> daemons within each partition
detect (by the lack of communication) a partitioning,
and then use the protocol to chose an active central manager.
As long as the partition remains,
and there exists an idle central manager within the partition,
there will be one active central manager within each partition.
When the network is repaired,
the protocol returns to having one central manager.
<P>
Through configuration,
a specific central manager machine may act as the
primary central manager.
While this machine is up and running,
it functions as the central manager.
After a failure of this primary central manager,
another idle central manager becomes the active one.
When the primary recovers,
it again becomes the central manager.
This is a recommended configuration,
if one of the central managers is a reliable machine,
which is expected to have very short periods of instability.
An alternative configuration allows the promoted
active central manager
(in the case that the central manager fails)
to stay active after the failed central manager machine
returns.
<P>
This high availability mechanism operates by monitoring
communication between machines.
Note that there is a significant difference in
communications between machines when
<OL>
<LI>a machine is down
</LI>
<LI>a specific daemon (the <SPAN CLASS="textit">condor_had</SPAN> daemon in this case)
is not running, yet the machine is functioning
</LI>
</OL>
The high availability mechanism distinguishes between these two,
and it operates based only on first
(when a central manager machine is down).
A lack of executing daemons does <SPAN CLASS="textit">not</SPAN>
cause the protocol to choose or use a new active central manager.
<P>
The central manager machine contains state information,
and this includes information about user priorities.
The information is kept in a single file,
and is used by the central manager machine.
Should the primary central manager fail,
a pool with high availability enabled would lose this information
(and continue operation, but with re-initialized priorities).
Therefore, the <SPAN CLASS="textit">condor_replication</SPAN> daemon exists
to replicate this file on all potential central manager machines.
This daemon promulgates the file in a way that is safe from error,
and more secure than dependence on a shared file system copy.
<P>
<A NAME="34368"></A>
<A NAME="34369"></A>
<A NAME="34306"></A>
<A NAME="34370"></A>
<A NAME="34371"></A>
<A NAME="34309"></A>
The <SPAN CLASS="textit">condor_replication</SPAN> daemon runs on each
potential central manager machine as well as on
the active central manager machine.
There is a unidirectional communication between the <SPAN CLASS="textit">condor_had</SPAN>
daemon and the <SPAN CLASS="textit">condor_replication</SPAN> daemon on each machine.
To properly do its job, the <SPAN CLASS="textit">condor_replication</SPAN> daemon must
transfer state files.
When it needs to transfer a file,
the <SPAN CLASS="textit">condor_replication</SPAN> daemons at both the sending and receiving
ends of the transfer invoke the <SPAN CLASS="textit">condor_transferer</SPAN> daemon.
These short lived daemons do the task of file transfer and then exit.
Do not place <TT>TRANSFERER</TT> into <TT>DAEMON_LIST</TT>,
as it is not a daemon that the <SPAN CLASS="textit">condor_master</SPAN> should invoke or
watch over.
<P>
<H3><A NAME="SECTION004112300000000000000"></A><A NAME="sec:HA-configuration"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">3</SPAN> Configuration
</H3>
<P>
The high availability of central manager machines is
enabled through configuration.
It is disabled by default.
All machines in a pool must be configured appropriately
in order to make the high availability mechanism work.
See
section <A HREF="3_3_Configuration.html#sec:HA-Config-File-Entries">3.3.29</A>,
for definitions of these configuration variables.
<P>
The stabilization period is the time it takes for the
<SPAN CLASS="textit">condor_had</SPAN> daemons
to detect a change in the pool state such as
an active central manager failure
or network partition, and recover from this change.
It may be computed using the following formula:
<PRE>
stabilization period = 12 * (number of central managers) *
$(HAD_CONNECTION_TIMEOUT)
</PRE>
<P>
To disable the high availability of central managers mechanism,
it is sufficient to remove
<TT>HAD</TT>, <TT>REPLICATION</TT>, and <TT>NEGOTIATOR</TT> from the
<TT>DAEMON_LIST</TT> configuration variable on all machines,
leaving only one <SPAN CLASS="textit">condor_negotiator</SPAN> in the pool.
<P>
To shut down a currently operating high availability mechanism,
follow the given steps.
All commands must be invoked
from a host which has administrative permissions
on all central managers.
The first three commands kill all <SPAN CLASS="textit">condor_had</SPAN>,
<SPAN CLASS="textit">condor_replication</SPAN>,
and all running <SPAN CLASS="textit">condor_negotiator</SPAN> daemons.
The last command is invoked on the host where the
single <SPAN CLASS="textit">condor_negotiator</SPAN> daemon is to run.
<P>
<OL>
<LI><code>condor_off -all -neg</code>
</LI>
<LI><code>condor_off -all -subsystem -replication</code>
</LI>
<LI><code>condor_off -all -subsystem -had</code>
</LI>
<LI><code>condor_on -neg</code>
</LI>
</OL>
<P>
When configuring <SPAN CLASS="textit">condor_had</SPAN> to control the <SPAN CLASS="textit">condor_negotiator</SPAN>,
if the default backoff constant value is too small, it can result in
a churning of the <SPAN CLASS="textit">condor_negotiator</SPAN>, especially in cases in which the
primary negotiator is unable to run due to misconfiguration.
In these cases, the <SPAN CLASS="textit">condor_master</SPAN> will kill the <SPAN CLASS="textit">condor_had</SPAN> after
the <SPAN CLASS="textit">condor_negotiator</SPAN> exists, wait a short period, then restart
<SPAN CLASS="textit">condor_had</SPAN>. The <SPAN CLASS="textit">condor_had</SPAN> will then win the election, so
the secondary <SPAN CLASS="textit">condor_negotiator</SPAN> will be killed, and the primary
will be restarted, only to exit again. If this happens too quickly,
neither <SPAN CLASS="textit">condor_negotiator</SPAN> will run long enough to complete a negotiation
cycle, resulting in no jobs getting started. Increasing this value
via <TT>MASTER_HAD_BACKOFF_CONSTANT</TT> <A NAME="34572"></A> <A NAME="34573"></A>
to be larger than a typical negotiation cycle can help solve this
problem.
<P>
To run a high availability pool without the replication feature,
do the following operations:
<OL>
<LI>Set the <TT>HAD_USE_REPLICATION</TT> <A NAME="34577"></A> <A NAME="34578"></A> configuration
variable to <TT>False</TT>,
and thus disable the replication on configuration level.
<P>
</LI>
<LI>Remove <TT>REPLICATION</TT> from both <TT>DAEMON_LIST</TT> and
<TT>DC_DAEMON_LIST</TT> in the configuration file.
</LI>
</OL>
<P>
<H3><A NAME="SECTION004112400000000000000"></A><A NAME="sec:HA-sample-config"></A>
<A NAME="34354"></A>
<BR>
<SPAN CLASS="arabic">3</SPAN>.<SPAN CLASS="arabic">11</SPAN>.<SPAN CLASS="arabic">2</SPAN>.<SPAN CLASS="arabic">4</SPAN> Sample Configuration
</H3>
This section provides sample configurations for high availability.
The two parts to this are the
configuration for the potential central manager machines,
and the configuration for the
machines within the pool that will <SPAN CLASS="textit">not</SPAN> be
central managers.
<P>
This is a sample configuration relating to the
high availability of central managers.
This is for the potential central manager machines.
<P>
<PRE>
##########################################################################
# A sample configuration file for central managers, to enable the #
# the high availability mechanism. #
##########################################################################
# unset this macro
CONDOR_HOST=
#########################################################################
## THE FOLLOWING MUST BE IDENTICAL ON ALL POTENTIAL CENTRAL MANAGERS. #
#########################################################################
## For simplicity in writing other expressions, define a variable
## for each potential central manager in the pool.
## These are samples.
CENTRAL_MANAGER1 = cm1.domain.name
CENTRAL_MANAGER2 = cm2.domain.name
## A list of all potential central managers in the pool.
COLLECTOR_HOST = $(CENTRAL_MANAGER1),$(CENTRAL_MANAGER2)
## Define the port number on which the condor_had daemon will
## listen. The port must match the port number used
## for when defining HAD_LIST. This port number is
## arbitrary; make sure that there is no port number collision
## with other applications.
HAD_PORT = 51450
HAD_ARGS = -p $(HAD_PORT)
## The following macro defines the port number condor_replication will listen
## on on this machine. This port should match the port number specified
## for that replication daemon in the REPLICATION_LIST
## Port number is arbitrary (make sure no collision with other applications)
## This is a sample port number
REPLICATION_PORT = 41450
REPLICATION_ARGS = -p $(REPLICATION_PORT)
## The following list must contain the same addresses
## as HAD_LIST. In addition, for each hostname, it should specify
## the port number of condor_replication daemon running on that host.
## This parameter is mandatory and has no default value
REPLICATION_LIST = \
$(CENTRAL_MANAGER1):$(REPLICATION_PORT), \
$(CENTRAL_MANAGER2):$(REPLICATION_PORT)
## The following list must contain the same addresses in the same order
## as COLLECTOR_HOST. In addition, for each hostname, it should specify
## the port number of condor_had daemon running on that host.
## The first machine in the list will be the PRIMARY central manager
## machine, in case HAD_USE_PRIMARY is set to true.
HAD_LIST = \
$(CENTRAL_MANAGER1):$(HAD_PORT), \
$(CENTRAL_MANAGER2):$(HAD_PORT)
## HAD connection time.
## Recommended value is 2 if the central managers are on the same subnet.
## Recommended value is 5 if Condor security is enabled.
## Recommended value is 10 if the network is very slow, or
## to reduce the sensitivity of HA daemons to network failures.
HAD_CONNECTION_TIMEOUT = 2
##If true, the first central manager in HAD_LIST is a primary.
HAD_USE_PRIMARY = true
##--------------------------------------------------------------------
## Host/IP access levels
##--------------------------------------------------------------------
## What machines have administrative rights for your pool? This
## defaults to your central manager. You should set it to the
## machine(s) where whoever is the condor administrator(s) works
## (assuming you trust all the users who log into that/those
## machine(s), since this is machine-wide access you're granting).
HOSTALLOW_ADMINISTRATOR = $(COLLECTOR_HOST)
## Negotiator access. Machines listed here are trusted central
## managers. You should normally not have to change this.
HOSTALLOW_NEGOTIATOR = $(COLLECTOR_HOST)
###################################################################
## THE PARAMETERS BELOW ARE ALLOWED TO BE DIFFERENT ON EACH #
## CENTRAL MANAGERS #
## THESE ARE MASTER SPECIFIC PARAMETERS
###################################################################
## The location of executable files
HAD = $(SBIN)/condor_had
REPLICATION = $(SBIN)/condor_replication
TRANSFERER = $(SBIN)/condor_transferer
## the master should start at least these five daemons
DAEMON_LIST = MASTER, COLLECTOR, NEGOTIATOR, HAD, REPLICATION
## DC_Daemon list should contain at least these five
DC_DAEMON_LIST = +HAD, REPLICATION
## Enables/disables the replication feature of HAD daemon
## Default: no
HAD_USE_REPLICATION = true
## Name of the file from the SPOOL directory that will be replicated
## Default: $(SPOOL)/Accountantnew.log
STATE_FILE = $(SPOOL)/Accountantnew.log
## Period of time between two successive awakenings of the replication daemon
## Default: 300
REPLICATION_INTERVAL = 300
# Period of time, in which transferer daemons have to accomplish the
## downloading/uploading process
## Default: 300
MAX_TRANSFERER_LIFETIME = 300
## Period of time between two successive sends of ClassAds to the collector by HAD
## Default: 300
HAD_UPDATE_INTERVAL = 300
## The HAD controls the negotiator, and should have a larger
## backoff constant
MASTER_NEGOTIATOR_CONTROLLER = HAD
MASTER_HAD_BACKOFF_CONSTANT = 360
## The size of the log file
MAX_HAD_LOG = 640000
## debug level
HAD_DEBUG = D_COMMAND
## location of the condor_had log file
HAD_LOG = $(LOG)/HADLog
## The size of replication log file
MAX_REPLICATION_LOG = 640000
## Replication debug level
REPLICATION_DEBUG = D_COMMAND
## Replication log file
REPLICATION_LOG = $(LOG)/ReplicationLog
## The size of transferer log file
MAX_TRANSFERER_LOG = 640000
## Replication debug level
TRANSFERER_DEBUG = D_COMMAND
## Replication log file
TRANSFERER_LOG = $(LOG)/TransferLog
</PRE>
<P>
Machines that are not potential central managers also
require configuration.
The following is a sample configuration relating to
high availability for machines that will <SPAN CLASS="textit">not</SPAN> be central managers.
<P>
<PRE>
##########################################################################
# Sample configuration relating to high availability for machines #
# that DO NOT run the condor_had daemon. #
##########################################################################
#unset this variable
CONDOR_HOST =
## For simplicity define a variable for each potential central manager
## in the pool.
CENTRAL_MANAGER1 = cm1.cs.technion.ac.il
CENTRAL_MANAGER2 = cm2.cs.technion.ac.il
## List of all potential central managers in the pool
COLLECTOR_HOST = $(CENTRAL_MANAGER1),$(CENTRAL_MANAGER2)
##--------------------------------------------------------------------
## Host/IP access levels
##--------------------------------------------------------------------
## Negotiator access. Machines listed here are trusted central
## managers. You should normally not need to change this.
HOSTALLOW_NEGOTIATOR = $(COLLECTOR_HOST)
## Now, with flocking (and HA) we need to let the SCHEDD trust the other
## negotiators we are flocking with as well. You should normally
## not need to change this.
HOSTALLOW_NEGOTIATOR_SCHEDD = $(COLLECTOR_HOST)
</PRE>
<P>
<P>
<DIV CLASS="navigation"><HR>
<!--Navigation Panel-->
<A NAME="tex2html1587"
HREF="3_12_Setting_Up.html">
<IMG WIDTH="37" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="next" SRC="next.png"></A>
<A NAME="tex2html1581"
HREF="3_Administrators_Manual.html">
<IMG WIDTH="26" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="up" SRC="up.png"></A>
<A NAME="tex2html1575"
HREF="3_10_Pool_Management.html">
<IMG WIDTH="63" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="previous" SRC="prev.png"></A>
<A NAME="tex2html1583"
HREF="Contents.html">
<IMG WIDTH="65" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="contents" SRC="contents.png"></A>
<A NAME="tex2html1585"
HREF="Index.html">
<IMG WIDTH="43" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="index" SRC="index.png"></A>
<BR>
<B> Next:</B> <A NAME="tex2html1588"
HREF="3_12_Setting_Up.html">3.12 Setting Up for</A>
<B> Up:</B> <A NAME="tex2html1582"
HREF="3_Administrators_Manual.html">3. Administrators' Manual</A>
<B> Previous:</B> <A NAME="tex2html1576"
HREF="3_10_Pool_Management.html">3.10 Pool Management</A>
<B> <A NAME="tex2html1584"
HREF="Contents.html">Contents</A></B>
<B> <A NAME="tex2html1586"
HREF="Index.html">Index</A></B> </DIV>
<!--End of Navigation Panel-->
<ADDRESS>
condor-admin@cs.wisc.edu
</ADDRESS>
</BODY>
</HTML>
|