-
Notifications
You must be signed in to change notification settings - Fork 183
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update OpenMPI to 4.1.0, and add extra support libraries #6523
Update OpenMPI to 4.1.0, and add extra support libraries #6523
Conversation
A new Pull Request was created by @fwyzard (Andrea Bocci) for branch IB/CMSSW_11_3_X/master. @cmsbuild, @smuzaffar, @mrodozov can you please review it and eventually sign? Thanks. |
Pull request #6523 was updated. |
2a202db
to
ba6fa0d
Compare
Pull request #6523 was updated. |
please test |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-d5cf7e/11860/summary.html Comparison Summary@slava77 comparisons for the following workflows were not done due to missing matrix map:
Summary:
|
ba6fa0d
to
39328c1
Compare
Pull request #6523 was updated. |
please test |
39328c1
to
96ceafe
Compare
Pull request #6523 was updated. |
please test |
Unified Communication X (UCX) provides an optimized communication layer for Message Passing (MPI), PGAS/OpenSHMEM libraries and RPC/data-centric applications. UCX utilizes high-speed networks for inter-node communication, and shared memory mechanisms for efficient intra-node communication. For more information see - https://www.openucx.org/ - https://github.com/openucx/ucx/tree/v1.9.0
See https://www.open-mpi.org/software/ompi/major-changes.php for major user-noticeable changes in the OpenMPI 4.1.x and 4.0.x series. See https://raw.githubusercontent.com/open-mpi/ompi/v4.1.x/NEWS for a detailed list of changes in the OpenMPI 4.1.x and 4.0.x series.
e4a3eea
to
143c147
Compare
Rebased to exclude the CUDA update from the changes. |
please test |
Pull request #6523 was updated. |
please test for slc7_aarch64_gcc9 |
please test for slc7_ppc64le_gcc9 |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-d5cf7e/12028/summary.html Comparison SummarySummary:
|
-1 Failed Tests: UnitTests AddOn Unit TestsI found errors in the following unit tests: ---> test testFWCoreConcurrency had ERRORS ---> test TestFWCoreServicesDriver had ERRORS ---> test TestRunnerFWCoreTFWLiteSelector had ERRORS ---> test TestConfigDP had ERRORS |
The workflows failing on
The unit tests failing on
All of them were already failing in the CMSSW_11_3_X_2021-01-08-2300 IB (see https://cmssdt.cern.ch/SDT/html/cmssdt-ib/#/ib/CMSSW_11_3_X). |
please test for slc7_aarch64_gcc9 |
-1 Failed Tests: UnitTests Unit TestsI found errors in the following unit tests: ---> test TestFWCoreServicesDriver had ERRORS ---> test TestFWCoreServicesDriver had ERRORS ---> test PhiTest had ERRORS ---> test TestConfigDP had ERRORS |
please test for slc7_ppc64le_gcc9 |
-1 Failed Tests: UnitTests RelVals Unit TestsI found errors in the following unit tests: ---> test test_PrepareInputDb had ERRORS ---> test test_PrepareInputDb had ERRORS ---> test test_MpsWorkFlow had ERRORS ---> test test_PrepareInputDb had ERRORS ---> test test_PrepareInputDb had ERRORS ---> test test_MpsWorkFlow had ERRORS ---> test testUnits had ERRORS ---> test testDD4hepCompactView had ERRORS ---> test testDD4hepFilteredViewGoTo had ERRORS ---> test testDD4hepFilteredViewFind had ERRORS ---> test testDD4hepFilteredView had ERRORS ---> test testDD4hepFilteredViewFirstChild had ERRORS ---> test testDD4hepFilteredViewGet had ERRORS ---> test testDD4hepFilteredViewLevel had ERRORS ---> test DetectorDescriptionDDCMSTestDriver had ERRORS and more ... RelVals
|
+externals |
This pull request is fully signed and it will be integrated in one of the next IB/CMSSW_11_3_X/master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @silviodonato, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2) |
Hi @fwyzard, this recipe calls for Question: should we change the recipe to use Thanks, |
Hi @iarspider good timing, I'm in the process of reviewing the build options for UCX and MPI to improve support for RDMA transfers, I'll add this to the list of things to check. |
Based on this comparison from a few years back, it looks like we should try to enable XPMEM: I'll have a look in parallel to enabling ROCm. |
Should be done in #7809 . |
Update OpenMPI to version 4.1.0
See https://www.open-mpi.org/software/ompi/major-changes.php for major user-noticeable changes in the OpenMPI 4.1.x and 4.0.x series.
See https://raw.githubusercontent.com/open-mpi/ompi/v4.1.x/NEWS for a detailed list of changes in the OpenMPI 4.1.x and 4.0.x series.
Add the UCX libraries version 1.9.0
Unified Communication X (UCX) provides an optimized communication layer for Message Passing (MPI), PGAS/OpenSHMEM libraries and RPC/data-centric applications.
UCX utilizes high-speed networks for inter-node communication, and shared memory mechanisms for efficient intra-node communication.
For more information see:
Add the NVIDIA GDRCopy library v2.1
GDRCopy is a low-latency GPU memory copy library based on NVIDIA GPUDirect RDMA technology.
For more details, see
Add the Portable Hardware Locality (hwloc) package
The Portable Hardware Locality (hwloc) software package provides a portable abstraction (across OS, versions, architectures, ...) of the hierarchical topology of modern architectures, including NUMA memory nodes, sockets, shared caches, cores and simultaneous multithreading.
It also gathers various system attributes such as cache and memory information as well as the locality of I/O devices such as network interfaces, InfiniBand HCAs or GPUs.
See https://www.open-mpi.org/projects/hwloc/ for more details.
Add numactl utilities and libnuma library version 2.0.14
Simple NUMA policy support. It consists of a numactl program to run other programs with a specific NUMA policy and a libnuma shared library ("NUMA API") to set NUMA policy in applications.
See https://github.com/numactl/numactl/blob/v2.0.14/README.md for more information.
Add libpciaccess, a generic PCI access library
libpciaccess is a generic PCI access library originally developed for the Xorg project; it is now used by various other projects, for example the hwloc library developed by the OpenMPI project.
The upstream git repository is at https://gitlab.freedesktop.org/xorg/lib/libpciaccess
To avoid the dependency on the "xorg-macros" autotool package, the CMSDIST package for libpciaccess is based on the Debian package (where the autogen step has already been performed) instead of the upstream git sources.