1
1
Граф коммитов

709 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
99c5afb397 Remove clang compiler warnings.
This commit was SVN r26523.
2012-05-29 23:36:06 +00:00
Ralph Castain
be6ed9c2df Allow partial use of allocations by specifying the max number of daemons (i.e., max VM size) for the job
This commit was SVN r26499.
2012-05-27 16:48:19 +00:00
Jeff Squyres
7969faf372 Fixes trac:3057: minor update to the man page to state that slot locations
in rankfiles use ''physical'' device indexes (vs. logical indexes).

This commit was SVN r26478.

The following Trac tickets were found above:
  Ticket 3057 --> https://svn.open-mpi.org/trac/ompi/ticket/3057
2012-05-23 11:43:33 +00:00
Ralph Castain
b217124bd8 Symlink instead of copy
This commit was SVN r26464.
2012-05-21 23:07:48 +00:00
Ralph Castain
da3873af6f Rename the mapreduce tool to "mr+" per the marketing types
This commit was SVN r26463.
2012-05-21 21:17:44 +00:00
Ralph Castain
a526afae92 Ensure we always cleanup local procs, no matter how we exited.
This commit was SVN r26454.
2012-05-18 23:37:40 +00:00
Ralph Castain
12ebc0e269 Don't need this to be a bin program as the class is captured in the jar
This commit was SVN r26453.
2012-05-18 23:37:18 +00:00
Ralph Castain
b16e43f489 Silence a warning on Mac
This commit was SVN r26449.
2012-05-18 15:27:04 +00:00
Ralph Castain
ca1b325738 Tweak the java setup so it works better on Mac. Only build mapreduce and allocators if hadoop support was requested.
This commit was SVN r26448.
2012-05-18 01:02:01 +00:00
Jeff Squyres
2d78728d38 Fix the macro name in the comment: it's EXTRA_DIST, not EXTRA_SOURCES.
This commit was SVN r26429.
2012-05-10 14:07:36 +00:00
Jeff Squyres
b325c17c72 It's a little weird to put in a blank _SOURCES line for the
HDFSFileFinder PROGRAM, but if we don't put in a _SOURCES line at all,
Automake will default to "HDFSFileFinter_class_SOURCES =
HDFSFileFinder.c", which clearly will cause problems.  

But we don't want to put the .java file in _SOURCES, either, because
we haven't configured Automake to handle Java (because current
versions of Automake only have GCJ, not other Java compilers).  So set
HDFSFileFinder_class_SOURCES to blank and list the .java file in
EXTRA_SOURCES (so that they get picked up for "make dist").

This commit was SVN r26424.
2012-05-10 13:54:51 +00:00
Ralph Castain
640f0610aa Fix the makefile to install the perl scripts properly
This commit was SVN r26416.
2012-05-09 14:06:02 +00:00
Ralph Castain
fd796cce0a Add an allocator tool for finding HDFS file locations and obtaining allocations for those nodes (supports both Hadoop 1 and 2). Split the Java support into two parts: detection of Java support and request for Java bindings.
This commit was SVN r26414.
2012-05-09 01:13:49 +00:00
Jeff Squyres
2ba10c37fe Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
   wholly replaced by hwloc.
 * Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
 * Update sm, smcuda, wv, and openib components to no longer use carto.
   Instead, use hwloc data.  There are still optimizations possible in
   the sm/smcuda BTLs (i.e., making multiple mpools).  Also, the old
   carto-based code found out how many NUMA nodes were ''available''
   -- not how many were used ''in this job''.  The new hwloc-using
   code computes the same value -- it was not updated to calculate how
   many NUMA nodes are used ''by this job.''
   * Note that I cannot compile the smcuda and wv BTLs -- I ''think''
     they're right, but they need to be verified by their owners.
 * The openib component now does a bunch of stuff to figure out where
   "near" OpenFabrics devices are.  '''THIS IS A CHANGE IN DEFAULT
   BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
   (I do not have a NUMA machine with an OpenFabrics device that is a
   non-uniform distance from multiple different NUMA nodes).
 * Completely rewrite the OMPI_Affinity_str() routine from the
   "affinity" mpiext extension.  This extension now understands
   hyperthreads; the output format of it has changed a bit to reflect
   this new information.
 * Bunches of minor changes around the code base to update names/types
   from maffinity/paffinity-based names to hwloc-based names.
 * Add some helper functions into the hwloc base, mainly having to do
   with the fact that we have the hwloc data reporting ''all''
   topology information, but sometimes you really only want the
   (online | available) data.

This commit was SVN r26391.
2012-05-07 14:52:54 +00:00
Ralph Castain
b2f77bf08f Extend the iof by adding two new components to support map-reduce IO chaining. Add a mapreduce tool for running such applications.
Fix the state machine to support multiple jobs being simultaneously launched as this is not only required for mapreduce, but can happen under comm-spawn applications as well.

This commit was SVN r26380.
2012-05-02 21:00:22 +00:00
Ralph Castain
a927318ea1 Add -N option as synonym for "npernode"
This commit was SVN r26367.
2012-05-01 16:18:14 +00:00
Jeff Squyres
501a86afe1 No need to include the generated files in the tarball. Thanks to
Eugene for pointing this out.

This commit was SVN r26339.
2012-04-25 14:19:18 +00:00
Ralph Castain
5d14fa7546 Fix mpi_abort, minimize error output.
This commit was SVN r26266.
2012-04-11 14:37:08 +00:00
Ralph Castain
14d5525fb1 Some minor cleanups. Get singletons working. Cleanup abort handling so it gets properly identified.
This commit was SVN r26261.
2012-04-10 19:08:54 +00:00
Ralph Castain
bd8b4f7f1e Sorry for mid-day commit, but I had promised on the call to do this upon my return.
Roll in the ORTE state machine. Remove last traces of opal_sos. Remove UTK epoch code.

Please see the various emails about the state machine change for details. I'll send something out later with more info on the new arch.

This commit was SVN r26242.
2012-04-06 14:23:13 +00:00
Ralph Castain
6dc44dc4b8 Look at the basename of the appname for the "java" keyword
This commit was SVN r26190.
2012-03-24 00:38:18 +00:00
Josh Hursey
a595525366 Add the callers name to the 'comm failed' error message, so we know between which two peers the communication failed.
This commit was SVN r26117.
2012-03-08 21:55:19 +00:00
Ralph Castain
c3cf46af65 Ensure install_dirs are filled in before parsing prefix
This commit was SVN r26093.
2012-03-03 23:14:15 +00:00
Ralph Castain
b8f093d1a0 Switch precedence - take the --prefix value over the absolute-path-to-mpirun so the backend prefix can be different from that of mpirun on hetero machines.
This commit was SVN r26085.
2012-03-02 22:59:13 +00:00
Ralph Castain
6c93dd13b0 Cleanup the prefix handling by mpirun. Important note: we do NOT support per-app_context prefixes!!
Don't let app_files trump given prefix values. Assign according to following precedence rules:

1. absolute path to mpirun, if given
2. --prefix value, if given to mpirun
3. default prefix, if configured with --enable-orterun-prefix-default
4. prefix from first app in app_file, if given
5. no prefix

This commit was SVN r26081.
2012-03-02 19:48:25 +00:00
Jeff Squyres
97b3603036 A bunch of fixes and improvements to Open MPI's various command line tools.
* fixed some bugs where "unknown" tokens were allowed on the command
   line (which should really only be used for ortertun).
 * if an unknown token is encountered, print a short error to stderr
   and quit with a nonzero exit status
 * if we don't find the right number of parameters to an option, print
   a short error to stderr and quit with a nonzero exit status
 * when --help is given, print the help message to stdout (not stderr)
   and quit with a zero exit status
 * added --showme:help option to the wrapper compilers
 * updated docs in opal/util/cmd_line.h
 * other small/miscellaneous CLI parsing bugs in various tools

I won't bore you with what we did before.  :-)  Here's some examples
of what the new behavior looks like:

{{{
% ompi_info --bogus
ompi_info: Error: unknown option "--bogus"
Type 'ompi_info --help' for usage.
% ompi_info --param bogus
ompi_info: Error: option "--param" did not have enough parameters (2)
Type 'ompi_info --help' for usage.
%
}}}

This commit was SVN r26072.
2012-02-29 17:52:38 +00:00
Ralph Castain
bc5886707f Document the mpirun exit status behavior
This commit was SVN r26009.
2012-02-22 23:47:00 +00:00
Ralph Castain
47c64ec837 Roll in Java bindings per telecon discussion. Man pages still under revision
This commit was SVN r25973.
2012-02-20 22:12:43 +00:00
Ralph Castain
d7d8a8cdf7 Some cleanup of the tmpdir session directory specifications. Remove the --tmpdir option from orterun as it was confusing. Create an orte_local_tmpdir_base mca param in its place. Clarify the role of the local vs remote vs global tmpdir base params, and ensure that you don't set conflicting options.
Remove the OMPI_PREFIX_ENV environmental variable as that was totally confusing as a way of setting a tmpdir base location.

This commit was SVN r25941.
2012-02-16 16:10:01 +00:00
Jeff Squyres
54cf60eb4b $(RM) is not a standard macro. Just use "rm" -- every platform has it.
This commit was SVN r25934.
2012-02-15 19:51:59 +00:00
Jeff Squyres
ae9503db6e Remove the sentence that says that --prefix is a per-context option.
This commit was SVN r25932.
2012-02-15 18:31:27 +00:00
Ralph Castain
61ac2bb11b If no session directories are being created, then we cannot create the debugger attachment fifo - so don't complain about it.
This commit was SVN r25802.
2012-01-27 04:05:23 +00:00
Ralph Castain
6db8c56cd4 Add local and node ranks to debugger daemon procs so the odls properly launches them
This commit was SVN r25774.
2012-01-25 03:17:10 +00:00
Ralph Castain
bf09133631 Correctly track the number of debugger daemons being spawned
This commit was SVN r25741.
2012-01-19 18:17:07 +00:00
Ralph Castain
6235a355de Correctly handle co-spawning of daemons when attaching to a running job. We cannot use the general process mappers as we only want debugger daemons spawned on nodes where application procs already exist. So custom build the map for the debugger daemon job, and have the plm just launch that job without doing its usual vm-spawn step.
This commit was SVN r25736.
2012-01-18 00:19:49 +00:00
Ralph Castain
fd0d9f73c6 Make preload_binaries an MCA param so it can be set in the default MCA parameters for a system
This commit was SVN r25728.
2012-01-17 17:16:05 +00:00
Shiqing Fan
f57f873404 Disable the debugger support for Windows.
This commit was SVN r25725.
2012-01-17 16:21:33 +00:00
Ralph Castain
ce7ddd0e10 Create the debugger attach fifo unless the user requests that we periodically poll insteaad.
This commit was SVN r25714.
2012-01-11 19:44:22 +00:00
Ralph Castain
bf103de66c My apologies for doing this outside of the usual time restrictions, but we need to get this in so we can make progress.
Move the ORTE-level debugger code back into orterun and out of the ORTE library to resolve symbol conflicts.

This commit was SVN r25713.
2012-01-11 15:53:09 +00:00
Jeff Squyres
a4c8bb27fa Pull in the MPIR_Breakpoint symbol via a dummy function in
debuggers_base_fns.c: orte_debugger_base_pull_mpir_breakpoint().

This commit was SVN r25660.
2011-12-15 18:39:34 +00:00
Nathan Hjelm
9dec101043 fix totalview launch through --debug
This commit was SVN r25654.
2011-12-15 15:19:13 +00:00
Ralph Castain
f531b09a8d Correctly handle -host and -hostfile options. Ensure the initial vm launch constrains itself to the union of specified hosts if those options are given. Get oversubscribe set correctly for that case.
This commit was SVN r25648.
2011-12-14 20:01:15 +00:00
Ralph Castain
7510339725 Remove stale orte_vm_launch param. Add a param that allows users to specify envars to forward/set so they can do it in the MCA param file instead of only via mpirun cmd line.
This commit was SVN r25580.
2011-12-06 21:31:22 +00:00
Ralph Castain
90b7f2a7bf The rest of the multi app_context fix. Remove the restriction on number of app_contexts that can have zero np specified as multiple mappers now support that use-case. Update the ranking algorithms to respect and track bookmarks. Ensure we properly set the oversubscribed flag on a per-node basis.
This commit was SVN r25578.
2011-12-06 17:28:29 +00:00
Ralph Castain
6fefe236a4 Warn users if they set opal_paffinity_alone, either to true or false, that this parameter is no longer functional - they must use the --bind-to option and its corresponding mca param.
This commit was SVN r25567.
2011-12-03 01:10:52 +00:00
Ralph Castain
c56acf60ca Although we never really thought about it, we made an unconscious assumption in the mapper system - we assumed that the daemons would be placed on nodes in the order that the nodes appear in the allocation. In other words, we assumed that the launch environment would map processes in node order.
Turns out, this isn't necessarily true. The Cray, for example, launches processes in a toroidal pattern, thus causing the daemons to wind up somewhere other than what we thought. Other environments (e.g., slurm) are also capable of such behavior, depending upon the default mapping algorithm they are told to use.

Resolve this problem by making the daemon-to-node assignment in the affected environments when the daemon calls back and tells us what node it is on. Order the nodes in the mapping list so they are in daemon-vpid order as opposed to the order in which they show in the allocation. For environments that don't exhibit this mapping behavior (e.g., rsh), this won't have any impact.

Also, clean up the vm launch procedure a little bit so it more closely aligns with the state machine implementation that is coming, and remove some lingering "slave" code.

This commit was SVN r25551.
2011-11-30 19:58:24 +00:00
Ralph Castain
b475421c16 As promised, rationalize the rsh support. Remove rshbase and the base rsh support, centralizing all rsh support into the rsh component. Remove the "slave" launch support as that experiment is complete. Fix tree spawn and make that the default method for rsh launch, turning it "off" for qrsh as that system does not support tree spawn.
This commit was SVN r25507.
2011-11-26 02:33:05 +00:00
Ralph Castain
9b59d8de6f This is actually a much smaller commit than it appears at first glance - it just touches a lot of files. The --without-rte-support configuration option has never really been implemented completely. The option caused various objects not to be defined and conditionally compiled some base functions, but did nothing to prevent build of the component libraries. Unfortunately, since many of those components use objects covered by the option, it caused builds to break if those components were allowed to build.
Brian dealt with this in the past by creating platform files and using "no-build" to block the components. This was clunky, but acceptable when only one organization was using that option. However, that number has now expanded to at least two more locations.

Accordingly, make --without-rte-support actually work by adding appropriate configury to prevent components from building when they shouldn't. While doing so, remove two frameworks (db and rmcast) that are no longer used as ORCM comes to a close (besides, they belonged in ORCM now anyway). Do some minor cleanups along the way.

This commit was SVN r25497.
2011-11-22 21:24:35 +00:00
Ralph Castain
6310361532 At long last, the fabled revision to the affinity system has arrived. A more detailed explanation of how this all works will be presented here:
https://svn.open-mpi.org/trac/ompi/wiki/ProcessPlacement

The wiki page is incomplete at the moment, but I hope to complete it over the next few days. I will provide updates on the devel list. As the wiki page states, the default and most commonly used options remain unchanged (except as noted below). New, esoteric and complex options have been added, but unless you are a true masochist, you are unlikely to use many of them beyond perhaps an initial curiosity-motivated experimentation.

In a nutshell, this commit revamps the map/rank/bind procedure to take into account topology info on the compute nodes. I have, for the most part, preserved the default behaviors, with three notable exceptions:

1. I have at long last bowed my head in submission to the system admin's of managed clusters. For years, they have complained about our default of allowing users to oversubscribe nodes - i.e., to run more processes on a node than allocated slots. Accordingly, I have modified the default behavior: if you are running off of hostfile/dash-host allocated nodes, then the default is to allow oversubscription. If you are running off of RM-allocated nodes, then the default is to NOT allow oversubscription. Flags to override these behaviors are provided, so this only affects the default behavior.

2. both cpus/rank and stride have been removed. The latter was demanded by those who didn't understand the purpose behind it - and I agreed as the users who requested it are no longer using it. The former was removed temporarily pending implementation.

3. vm launch is now the sole method for starting OMPI. It was just too darned hard to maintain multiple launch procedures - maybe someday, provided someone can demonstrate a reason to do so.

As Jeff stated, it is impossible to fully test a change of this size. I have tested it on Linux and Mac, covering all the default and simple options, singletons, and comm_spawn. That said, I'm sure others will find problems, so I'll be watching MTT results until this stabilizes.

This commit was SVN r25476.
2011-11-15 03:40:11 +00:00
Ralph Castain
729935dffb Minor cleanups, mirroring what Jeff did to ompi_info
This commit was SVN r25438.
2011-11-05 00:42:49 +00:00