Merge pull request #1727 from jsquyres/pr/mpirun-timeout-and-friends

mpirun.1in: add descriptions of new options

contrib/completion/mpirun.zsh

@@ -158,6 +158,7 @@ _mpirun() {
 	'(-do-not-launch --do-not-launch)'{-do-not-launch,--do-not-launch}'[Perform all necessary operations to prepare to launch the application, but do not actually launch it]' \
 	'(-do-not-resolve --do-not-resolve)'{-do-not-resolve,--do-not-resolve}'[Do not attempt to resolve interfaces]' \
 	'(-enable-recovery --enable-recovery)'{-enable-recovery,--enable-recovery}'[Enable recovery from process failure (default: disabled)]' \
+	'(-get-stack-traces --get-stack-traces)'{-get-stack-traces,--get-stack-traces}'[Upon timeout, obtain stack traces from all still-alive MPI processes (default: disabled)]' \
 	'*'{-gmca,--gmca}'[Pass global MCA parameters that are applicable to all contexts (arg0 is the parameter name; arg1 is the parameter value)]:mca variable name:->mca_variable_name:mca variable value:->mca_variable_value' \
 	'(- *)'{-h,--help}'[Help message]' \
 	'*'{-H,-host,--host}'[List of hosts to invoke processes on]:hostnames:' \
@@ -192,6 +193,7 @@ _mpirun() {
 	'(-report-child-jobs-separately --report-child-jobs-separately)'{-report-child-jobs-separately,--report-child-jobs-separately}'[Return the exit status of the primary job only]' \
 	'(-report-events --report-events)'{-report-events,--report-events}'[Report events to a tool listening at the specified URI]:URI:' \
 	'(-report-pid --report-pid)'{-report-pid,--report-pid}'[Printout pid on stdout (-), stderr (+), or a file (anything else)]:report file:_report_file' \
+	'(-report-state-upon-timeout --report-state-upon-timeout)'{-report-state-upon-timeout,--report-state-upon-timeout}'[Upon timeout, print run-time status of each process]' \
 	'(-report-uri --report-uri)'{-report-uri,--report-uri}'[Printout URI on stdout (-), stderr (+), or a file (anything else)]:report file:_report_file' \
 	'(-rf --rankfile)'{-rf,--rankfile}'[Provide a rankfile file]:rank file:_files' \
 	'(-s --preload-binary)'{-s,--preload-binary}'[Preload the binary on the remote machine before starting the remote process.]' \
@@ -202,6 +204,7 @@ _mpirun() {
 	'(-staged --staged)'{-staged,--staged}'[Used staged execution if inadequate resources are present (cannot support MPI jobs)]' \
 	'(-stdin --stdin)'{-stdin,--stdin}'[Specify procs to receive stdin \[rank, all, none\] (default: 0, indicating rank 0)]:rank list:' \
 	'(-tag-output --tag-output)'{-tag-output,--tag-output}'[Tag all output with \[job,rank\]]' \
+	'(-timeout --timeout)'{-timeout,--timeout}'[Timeout, in seconds, for the entire job]' \
 	'(-timestamp-output --timestamp-output)'{-timestamp-output,--timestamp-output}'[Timestamp all application process output]' \
 	'(-use-hwthread-cpus --use-hwthread-cpus)'{-use-hwthread-cpus,--use-hwthread-cpus}'[Use hardware threads as independent cpus]' \
 	'(-use-regexp --use-regexp)'{-use-regexp,--use-regexp}'[Use regular expressions for launch]' \

orte/tools/orterun/help-orterun.txt

@@ -642,14 +642,13 @@ number of processes to run:
 Please correct this value and try again.
 #
 [orterun:timeout]
-The user-provided time limit for job execution has been
-reached:
+The user-provided time limit for job execution has been reached:
 
-  MPIEXEC_TIMEOUT: %d seconds
+  Timeout: %d seconds
 
 The job will now be aborted.  Please check your code and/or
-adjust/remove the job execution time limit (as specified by
-MPIEXEC_TIMEOUT in your environment or --timeout on the command line).
+adjust/remove the job execution time limit (as specified by --timeout
+command line option or MPIEXEC_TIMEOUT environment variable).
 #
 [orterun:conflict-env-set]
 ERROR: You have attempted to pass environment variables to Open MPI

orte/tools/orterun/orterun.1in

@@ -1,5 +1,5 @@
 .\" -*- nroff -*-
-.\" Copyright (c) 2009-2014 Cisco Systems, Inc.  All rights reserved.
+.\" Copyright (c) 2009-2016 Cisco Systems, Inc.  All rights reserved.
 .\" Copyright (c) 2008-2009 Sun Microsystems, Inc.  All rights reserved.
 .\" $COPYRIGHT$
 .\"
@@ -529,12 +529,41 @@ MCA parameter.
 .
 .
 .TP
+.B --get-stack-traces
+When paired with the
+.B --timeout
+option,
+.I mpirun
+will obtain and print out stack traces from all launched processes
+that are still alive when the timeout expires.  Note that obtaining
+stack traces can take a little time and produce a lot of output,
+especially for large process-count jobs.
+.
+.
+.TP
 .B -debugger\fR,\fP --debugger
 Sequence of debuggers to search for when \fI--debug\fP is used (i.e.
 a synonym for \fIorte_base_user_debugger\fP MCA parameter).
 .
 .
 .TP
+.B --timeout \fR<seconds>
+The maximum number of seconds that
+.I mpirun
+(also known as
+.I mpiexec\fR,\fI oshrun\fR,\fI orterun\fR,\fI
+etc.)
+will run.  After this many seconds,
+.I mpirun
+will abort the launched job and exit with a non-zero exit status.
+Using
+.B --timeout
+can be also useful when combined with the
+.B --get-stack-traces
+option.
+.
+.
+.TP
 .B -tv\fR,\fP --tv
 Launch processes under the TotalView debugger.
 Deprecated backwards compatibility flag. Synonym for \fI--debug\fP.
@@ -661,6 +690,14 @@ without clutter from mpirun itself.
 Disable the automatic --prefix behavior
 .
 .
+.TP
+.B --report-state-on-timeout
+When paired with the
+.B --timeout
+command line option, report the run-time subsystem state of each
+process when the timeout expires.
+.
+.
 .P
 There may be other options listed with \fImpirun --help\fP.
 .
@@ -669,12 +706,9 @@ There may be other options listed with \fImpirun --help\fP.
 .
 .TP
 .B MPIEXEC_TIMEOUT
-The maximum number of seconds that
-.I mpirun
-.RI ( mpiexec )
-will run.  After this many seconds,
-.I mpirun
-will abort the launched job and exit.
+Synonym for the
+.B --timeout
+command line option.
 .
 .
 .\" **************************
@@ -1541,6 +1575,19 @@ In the event that one or more processes exit before calling MPI_FINALIZE, the
 return value of the MPI_COMM_WORLD rank of the process that \fImpirun\fP first notices died
 before calling MPI_FINALIZE will be returned.  Note that, in general, this will
 be the first process that died but is not guaranteed to be so.
+.
+.PP
+If the
+.B --timeout
+command line option is used and the timeout expires before the job
+completes (thereby forcing
+.I mpirun
+to kill the job)
+.I mpirun
+will return an exit status equivalent to the value of
+.B ETIMEDOUT
+(which is typically 110 on Linux and OS X systems).
+
 .
 .\" **************************
 .\"    See Also Section