percona-toolkit 2.1.2-1 / usr / bin / pt-pmp

#!/usr/bin/env bash

# This program is part of Percona Toolkit: http://www.percona.com/software/
# See "COPYRIGHT, LICENSE, AND WARRANTY" at the end of this file for legal
# notices and disclaimers.

TOOL="pt-pmp"

# ###########################################################################
# tmpdir package
# This package is a copy without comments from the original.  The original
# with comments and its test file can be found in the Bazaar repository at,
#   lib/bash/tmpdir.sh
#   t/lib/bash/tmpdir.sh
# See https://launchpad.net/percona-toolkit for more information.
# ###########################################################################


set -u

PT_TMPDIR=""

mk_tmpdir() {
   local dir="${1:-""}"

   if [ -n "$dir" ]; then
      if [ ! -d "$dir" ]; then
         mkdir "$dir" || die "Cannot make tmpdir $dir"
      fi
      PT_TMPDIR="$dir"
   else
      local tool="${0##*/}"
      local pid="$$"
      PT_TMPDIR=`mktemp -d -t "${tool}.${pid}.XXXXXX"` \
         || die "Cannot make secure tmpdir"
   fi
}

rm_tmpdir() {
   if [ -n "$PT_TMPDIR" ] && [ -d "$PT_TMPDIR" ]; then
      rm -rf "$PT_TMPDIR"
   fi
   PT_TMPDIR=""
}

# ###########################################################################
# End tmpdir package
# ###########################################################################

set +u

usage() {
   if [ "${OPT_ERR}" ]; then
      echo "${OPT_ERR}" >&2
   fi
   echo "Usage: pt-pmp [OPTIONS] [FILES]" >&2
   echo "For more information, 'man pt-pmp' or 'perldoc $0'" >&2
   exit 1
}

# Actually does the aggregation.  The arguments are the max number of functions
# to aggregate, and the files to read.  If maxlen=0, it means infinity.  We have
# to pass the maxlen argument into this function to make maxlen testable.
aggregate_stacktrace() {
   local maxlen=$1;
   shift
   awk "
      BEGIN {
         s = \"\";
      }
      /^Thread/ {
         if ( s != \"\" ) {
            print s;
         }
         s = \"\";
         c = 0;
      }
      /^\#/ {
         if ( \$2 ~ /0x/ ) {
            if ( \$4 ~/void|const/ ) {
               targ = \$5;
            }
            else {
               targ = \$4;
            }
            if ( targ ~ /[<\\(]/ ) {
               targ = substr(\$0, index(\$0, \" in \") + 4);
               if ( targ ~ / from / ) {
                  targ = substr(targ, 1, index(targ, \" from \") - 1);
               }
               if ( targ ~ / at / ) {
                  targ = substr(targ, 1, index(targ, \" at \") - 1);
               }
               # Shorten C++ templates, e.g. in t/samples/stacktrace-004.txt
               while ( targ ~ />::/ ) {
                  if ( 0 == gsub(/<[^<>]*>/, \"\", targ) ) {
                     break;
                  }
               }
               # Further shorten argument lists.
               while ( targ ~ /\\(/ ) {
                  if ( 0 == gsub(/\\([^()]*\\)/, \"\", targ) ) {
                     break;
                  }
               }
               # Remove void and const decorators.
               gsub(/ ?(void|const) ?/, \"\", targ);
               gsub(/ /, \"\", targ);
            }
            else if ( targ ~ /\\?\\?/ && \$2 ~ /[1-9]/ ) {
               # Substitute ?? by the name of the library.
               targ = \$NF;
               while ( targ ~ /\\// ) {
                  targ = substr(targ, index(targ, \"/\") + 1);
               }
               targ = substr(targ, 1, index(targ, \".\") - 1);
               targ = targ \"::??\";
            }
         }
         else {
            targ = \$2;
         }
         # get rid of long symbol names such as 'pthread_cond_wait@@GLIBC_2.3.2'
         if ( targ ~ /@@/ ) {
            fname = substr(targ, 1, index(targ, \"@@\") - 1);
         }
         else {
            fname = targ;
         }
         if ( ${maxlen:-0} == 0 || c < ${maxlen:-0} ) {
            if (s != \"\" ) {
               s = s \",\" fname;
            }
            else {
               s = fname;
            }
         }
         c++;
      }
      END {
         print s
      }
   " "$@" | sort | uniq -c | sort -r -n -k 1,1
}

# The main program to run.
main() {

   # Get command-line options
   for o; do
      case "${o}" in
         --)
            shift; break;
            ;;
         --help)
            usage;
            ;;
         -b)
            shift; OPT_b="${1}"; shift;
            ;;
         -i)
            shift; OPT_i="${1}"; shift;
            ;;
         -k)
            shift; OPT_k="${1}"; shift;
            ;;
         -l)
            shift; OPT_l="${1}"; shift;
            ;;
         -p)
            shift; OPT_p="${1}"; shift;
            ;;
         -s)
            shift; OPT_s="${1}"; shift;
            ;;
         -*)
            OPT_ERR="Unknown option ${o}."
            usage
            ;;
      esac
   done
   export OPT_i="${OPT_i:-1}";
   export OPT_k="${OPT_k:-}";
   export OPT_l="${OPT_l:-0}";
   export OPT_b="${OPT_b:-mysqld}";
   export OPT_p="${OPT_p:-}";
   export OPT_s="${OPT_s:-0}";

   if [ -z "${1}" ]; then
      # There's no file to analyze, so we'll make one.
      if [ -z "${OPT_p}" ]; then
         OPT_p=$(pidof -s "${OPT_b}" 2>/dev/null);
         if [ -z "${OPT_p}" ]; then
            OPT_p=$(pgrep -o -x "${OPT_b}" 2>/dev/null)
         fi
         if [ -z "${OPT_p}" ]; then
            OPT_p=$(ps -eaf | grep "${OPT_b}" | grep -v grep | awk '{print $2}' | head -n1);
         fi
      fi
      date;
      for x in $(seq 1 $OPT_i); do
         gdb -ex "set pagination 0" -ex "thread apply all bt" -batch -p $OPT_p >> "${OPT_k:-$PT_TMPDIR/percona-toolkit}"
         date +'TS %N.%s %F %T' >> "${OPT_k:-$PT_TMPDIR/percona-toolkit}"
         sleep $OPT_s
      done
   fi

   if [ $# -eq 0 ]; then
      aggregate_stacktrace "${OPT_l}" "${OPT_k:-$PT_TMPDIR/percona-toolkit}"
   else
      aggregate_stacktrace "${OPT_l}" "$@"
   fi
}

# Execute the program if it was not included from another file.  This makes it
# possible to include without executing, and thus test.
if    [ "${0##*/}" = "$TOOL" ] \
   || [ "${0##*/}" = "bash" -a "$_" = "$0" ]; then
   mk_tmpdir
   main "$@"
   rm_tmpdir
fi

# ############################################################################
# Documentation
# ############################################################################
:<<'DOCUMENTATION'
=pod

=head1 NAME

pt-pmp - Aggregate GDB stack traces for a selected program.

=head1 SYNOPSIS

Usage: pt-pmp [OPTIONS] [FILES]

pt-pmp is a poor man's profiler, inspired by L<http://poormansprofiler.org>.
It can create and summarize full stack traces of processes on Linux.
Summaries of stack traces can be an invaluable tool for diagnosing what
a process is waiting for.

=head1 RISKS

The following section is included to inform users about the potential risks,
whether known or unknown, of using this tool.  The two main categories of risks
are those created by the nature of the tool (e.g. read-only tools vs. read-write
tools) and those created by bugs.

pt-pmp is a read-only tool.  However, collecting GDB stacktraces is achieved by
attaching GDB to the program and printing stack traces from all threads. This
will freeze the program for some period of time, ranging from a second or so to
much longer on very busy systems with a lot of memory and many threads in the
program.  In the tool's default usage as a MySQL profiling tool, this means that
MySQL will be unresponsive while the tool runs, although if you are using the
tool to diagnose an unresponsive server, there is really no reason not to do
this.  In addition to freezing the server, there is also some risk of the server
crashing or performing badly after GDB detaches from it.

At the time of this release, we know of no bugs that could cause serious harm
to users.

The authoritative source for updated information is always the online issue
tracking system.  Issues that affect this tool will be marked as such.  You can
see a list of such issues at the following URL:
L<http://www.percona.com/bugs/pt-pmp>.

See also L<"BUGS"> for more information on filing bugs and getting help.

=head1 DESCRIPTION

pt-pmp performs two tasks: it gets a stack trace, and it summarizes the stack
trace.  If a file is given on the command line, the tool skips the first step
and just aggregates the file.

To summarize the stack trace, the tool extracts the function name (symbol)
from each level of the stack, and combines them with commas.  It does this
for each thread in the output.  Afterwards, it sorts similar threads together
and counts how many of each one there are, then sorts them most-frequent first.

=head1 OPTIONS

Options must precede files on the command line.

=over

=item -b BINARY

Which binary to trace (default mysqld)

=item -i ITERATIONS

How many traces to gather and aggregate (default 1)

=item -k KEEPFILE

Keep the raw traces in this file after aggregation

=item -l NUMBER

Aggregate only first NUMBER functions; 0=infinity (default 0)

=item -p PID

Process ID of the process to trace; overrides -b

=item -s SLEEPTIME

Number of seconds to sleep between iterations (default 0)

=back

=head1 ENVIRONMENT

This tool does not use any environment variables.

=head1 SYSTEM REQUIREMENTS

This tool requires Bash v3 or newer.

=head1 BUGS

For a list of known bugs, see L<http://www.percona.com/bugs/pt-pmp>.

Please report bugs at L<https://bugs.launchpad.net/percona-toolkit>.
Include the following information in your bug report:

=over

=item * Complete command-line used to run the tool

=item * Tool L<"--version">

=item * MySQL version of all servers involved

=item * Output from the tool including STDERR

=item * Input files (log/dump/config files, etc.)

=back

If possible, include debugging output by running the tool with C<PTDEBUG>;
see L<"ENVIRONMENT">.

=head1 DOWNLOADING

Visit L<http://www.percona.com/software/percona-toolkit/> to download the
latest release of Percona Toolkit.  Or, get the latest release from the
command line:

   wget percona.com/get/percona-toolkit.tar.gz

   wget percona.com/get/percona-toolkit.rpm

   wget percona.com/get/percona-toolkit.deb

You can also get individual tools from the latest release:

   wget percona.com/get/TOOL

Replace C<TOOL> with the name of any tool.

=head1 AUTHORS

Baron Schwartz, based on a script by Domas Mituzas (L<http://poormansprofiler.org/>)

=head1 ABOUT PERCONA TOOLKIT

This tool is part of Percona Toolkit, a collection of advanced command-line
tools developed by Percona for MySQL support and consulting.  Percona Toolkit
was forked from two projects in June, 2011: Maatkit and Aspersa.  Those
projects were created by Baron Schwartz and developed primarily by him and
Daniel Nichter, both of whom are employed by Percona.  Visit
L<http://www.percona.com/software/> for more software developed by Percona.

=head1 COPYRIGHT, LICENSE, AND WARRANTY

This program is copyright 2010-2011 Baron Schwartz, 2011-2012 Percona Inc.
Feedback and improvements are welcome.

THIS PROGRAM IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED
WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

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, version 2; OR the Perl Artistic License.  On UNIX and similar
systems, you can issue `man perlgpl' or `man perlartistic' to read these
licenses.

You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 59 Temple
Place, Suite 330, Boston, MA  02111-1307  USA.

=head1 VERSION

pt-pmp 2.1.2

=cut

DOCUMENTATION