WHAT: Merge the PMIx branch into the devel repo, creating a new
OPAL “lmix” framework to abstract PMI support for all RTEs.
Replace the ORTE daemon-level collectives with a new PMIx
server and update the ORTE grpcomm framework to support
server-to-server collectives
WHY: We’ve had problems dealing with variations in PMI implementations,
and need to extend the existing PMI definitions to meet exascale
requirements.
WHEN: Mon, Aug 25
WHERE: https://github.com/rhc54/ompi-svn-mirror.git
Several community members have been working on a refactoring of the current PMI support within OMPI. Although the APIs are common, Slurm and Cray implement a different range of capabilities, and package them differently. For example, Cray provides an integrated PMI-1/2 library, while Slurm separates the two and requires the user to specify the one to be used at runtime. In addition, several bugs in the Slurm implementations have caused problems requiring extra coding.
All this has led to a slew of #if’s in the PMI code and bugs when the corner-case logic for one implementation accidentally traps the other. Extending this support to other implementations would have increased this complexity to an unacceptable level.
Accordingly, we have:
* created a new OPAL “pmix” framework to abstract the PMI support, with separate components for Cray, Slurm PMI-1, and Slurm PMI-2 implementations.
* Replaced the current ORTE grpcomm daemon-based collective operation with an integrated PMIx server, and updated the grpcomm APIs to provide more flexible, multi-algorithm support for collective operations. At this time, only the xcast and allgather operations are supported.
* Replaced the current global collective id with a signature based on the names of the participating procs. The allows an unlimited number of collectives to be executed by any group of processes, subject to the requirement that only one collective can be active at a time for a unique combination of procs. Note that a proc can be involved in any number of simultaneous collectives - it is the specific combination of procs that is subject to the constraint
* removed the prior OMPI/OPAL modex code
* added new macros for executing modex send/recv to simplify use of the new APIs. The send macros allow the caller to specify whether or not the BTL supports async modex operations - if so, then the non-blocking “fence” operation is used, if the active PMIx component supports it. Otherwise, the default is a full blocking modex exchange as we currently perform.
* retained the current flag that directs us to use a blocking fence operation, but only to retrieve data upon demand
This commit was SVN r32570.
As part of reviewing CMR #4860, I talked through r32517 with Ralph.
In attempt to fix various rsh quoting problems, r32517 removed all the
quoting from the main code path and then only added it back in at the
end in some cases.
This commit puts back the quoting parts that were removed in r32517
(r32517 fixed 2 other important bugs: a) change "--<foo>" to "--mca
<foo_equivalent> 1" so that de-duplication works, and b) change a !=
to ==).
refs trac:4860
This commit was SVN r32524.
The following SVN revision numbers were found above:
r32517 --> open-mpi/ompi@7342bce58f
The following Trac tickets were found above:
Ticket 4860 --> https://svn.open-mpi.org/trac/ompi/ticket/4860
Also discovered that the rsh launcher is not picking up --enable-orterun-prefix-by-default when invoked during singleton comm_spawn, but I was unable to see why that was happening and ran out of time.
cmr=v1.8.2:reviewer=rhc
This commit was SVN r32229.
We have been getting several requests for new collectives that need to be inserted in various places of the MPI layer, all in support of either checkpoint/restart or various research efforts. Until now, this would require that the collective id's be generated at launch. which required modification
s to ORTE and other places. We chose not to make collectives reusable as the race conditions associated with resetting collective counters are daunti
ng.
This commit extends the collective system to allow self-generation of collective id's that the daemons need to support, thereby allowing developers to request any number of collectives for their work. There is one restriction: RTE collectives must occur at the process level - i.e., we don't curren
tly have a way of tagging the collective to a specific thread. From the comment in the code:
* In order to allow scalable
* generation of collective id's, they are formed as:
*
* top 32-bits are the jobid of the procs involved in
* the collective. For collectives across multiple jobs
* (e.g., in a connect_accept), the daemon jobid will
* be used as the id will be issued by mpirun. This
* won't cause problems because daemons don't use the
* collective_id
*
* bottom 32-bits are a rolling counter that recycles
* when the max is hit. The daemon will cleanup each
* collective upon completion, so this means a job can
* never have more than 2**32 collectives going on at
* a time. If someone needs more than that - they've got
* a problem.
*
* Note that this means (for now) that RTE-level collectives
* cannot be done by individual threads - they must be
* done at the overall process level. This is required as
* there is no guaranteed ordering for the collective id's,
* and all the participants must agree on the id of the
* collective they are executing. So if thread A on one
* process asks for a collective id before thread B does,
* but B asks before A on another process, the collectives will
* be mixed and not result in the expected behavior. We may
* find a way to relax this requirement in the future by
* adding a thread context id to the jobid field (maybe taking the
* lower 16-bits of that field).
This commit includes a test program (orte/test/mpi/coll_test.c) that cycles 100 times across barrier and modex collectives.
This commit was SVN r32203.
The alps ras and plm components were broken by recent changes in ORTE. This
commit resolves those issues.
Changes:
- Define PMI2_SUCCESS if it isn't defined. This fixes a problem with Cray's
PMI implementation which does not define (for some reason) PMI2_SUCCESS. We
had previously just used PMI_SUCCESS.
- Add missing definition and a typo in pml_alps_module.
- launch_id is no longer available in the orte_node_t structure. Use the
attribute lookup to get the value.
- Do not use an O(n^2) sorting algorithm when putting alps nodes in order. Use
opal_list_sort instead (O(nlogn)).
This commit was SVN r32076.
http://www.open-mpi.org/community/lists/devel/2014/05/14822.php
Revamp the ORTE global data structures to reduce memory footprint and add new features. Add ability to control/set cpu frequency, though this can only be done if the sys admin has setup the system to support it (or you run as root).
This commit was SVN r31916.
This commit is a slightly better workaround to prevent mesages of
the form:
[unset]:_pmi_alps_get_apid:alps_app_lli_put_request failed
[unset]:_pmi_alps_get_appLayout:pmi_alps_get_apid returned with error: Bad file descriptor
It works by completely disabling PMI in the application process when using
mpirun. This should not be an issue for any apps.
cmr=v1.8.2:reviewer=rhc
This commit was SVN r31882.
The HNP can't know the precise reason, of course - all it knows is that the daemon failed. So output a generic error message that provides guidance on probable causes.
Refs trac:4571
This commit was SVN r31589.
The following Trac tickets were found above:
Ticket 4571 --> https://svn.open-mpi.org/trac/ompi/ticket/4571
Prior to r29058, this same logic was in place (i.e., ensure that the
extra fd to /dev/null is closed). It looks like it was accidentally
removed in the ORTE conversion to the state machine in r29058.
This ''might'' have something to do with many hangs that we're seeing
in Cisco MTT with jobs that exhibit failure (e.g., call MPI_ABORT)...?
cmr=v1.8.2:reviewer=rhc
This commit was SVN r31469.
The following SVN revision numbers were found above:
r29058 --> open-mpi/ompi@a200e4f865
newer
This commit adds a workaround for messages printed by the Cray PMI library
when launching using mpirun. We are still talking with Cray to find a
better fix but this will silence the warnings for now.
cmr=v1.8.1:reviewer=manjugv
This commit was SVN r31352.
See the ticket for more details.
cmr=v1.8.1:reviewer=rhc:ticket=4489
This commit was SVN r31351.
The following Trac tickets were found above:
Ticket 4489 --> https://svn.open-mpi.org/trac/ompi/ticket/4489
If we are aborting, then set the flags so the HNP directly sends an exit command to each daemon. Make it the halt_vm command so the remote daemon doesn't try to relay it, but instead just exits without waiting for its routed children to exit first.
cmr=v1.8.1:reviewer=jsquyres:subject=fix hangs due to abort prior to daemon wireup
This commit was SVN r31304.
The problem arises when a hostfile is used, and the user provides host names without specifying the slots= paramater. In these cases, we assign slots=1, but automatically allow oversubscription since that number isn't confirmed. We then provide a separate parameter by which the user can direct that we assign the number of slots based on the sensed hardware - e.g., by telling us to set the #slots equal to the #cores on each node. However, this has been set to "off" by default.
In order to make this a little less complex for the user, set the default such that we automatically set #slots equal to #cores (or #hwt's if use_hwthreads_as_cpus has been set) only for those cases where the user provides names in a hostfile but does not provide slot information.
Also cleanup some a couple of issues in the mapping/binding system:
* ensure we only override the binding directive if we are oversubscribed *and* overload is not allowed
* ensure that the MPI procs don't attempt to bind themselves if they are launched by an orted as any binding directive (no matter what it was) would have been serviced by the orted on launch
* minor cleanup to the warning message when oversubscribed and binding was requested
cmr=v1.7.5:reviewer=rhc:subject=update mapping/binding system
This commit was SVN r30909.
