3.9. Network Support
3.9.1. Main network support models
There are multiple MPI network models available in this release:
ob1supports a variety of networks using BTL (“Byte Transfer Layer”) plugins that can be used in combination with each other:
self: Loopback (send-to-self)
sm: Shared memory, including single-copy technologies: XPMEM, Linux CMA, as Linux KNEM, as well as traditional copy-in/copy-out shared memory.
smcuda: CUDA-enabled shared memory
usnic: Cisco usNIC
ugni: uGNI (Cray Gemini, Aries)
cmsupports a smaller number of networks (and they cannot be used together), but may provide better overall MPI performance by utilizing MTL (“Matching Transport Layer”) plugins:
OpenFabrics Interfaces (“libfabric” tag matching)
Intel Omni-Path PSM2 (version 11.2.173 or later)
Intel True Scale PSM (QLogic InfiniPath)
ucxuses the Unified Communication X (UCX) communication library. This is an open-source project developed in collaboration between industry, laboratories, and academia to create an open-source production grade communication framework for data centric and high-performance applications. The UCX library can be downloaded from repositories (e.g., Fedora/RedHat yum repositories). The UCX library is also part of Mellanox OFED and Mellanox HPC-X binary distributions.
UCX currently supports:
OpenFabrics Verbs (including InfiniBand and RoCE)
NVIDIA CUDA drivers
While users can manually select any of the above transports at run time, Open MPI will select a default transport as follows:
If InfiniBand devices are available, use the UCX PML.
If PSM, PSM2, or other tag-matching-supporting Libfabric transport devices are available (e.g., Cray uGNI), use the
cmPML and a single appropriate corresponding
Otherwise, use the
ob1PML and one or more appropriate
Users can override Open MPI’s default selection algorithms and force
the use of a specific transport if desired by setting the
parameter (and potentially the
mtl MCA parameters) at
shell$ mpirun --mca pml ob1 --mca btl [comma-delimted-BTLs] ...
shell$ mpirun --mca pml cm --mca mtl [MTL] ...
shell$ mpirun --mca pml ucx ...
There is a known issue when using UCX with very old Mellanox Infiniband HCAs, in particular HCAs preceding the introduction of the ConnectX product line, which can result in Open MPI crashing in MPI_Finalize. This issue is addressed by UCX release 1.9.0 and newer.
3.9.2. Miscellaneous network notes
The main OpenSHMEM network model is
ucx; it interfaces directly with UCX.
In prior versions of Open MPI, InfiniBand and RoCE support was provided through the
ob1PML plugins. Starting with Open MPI 4.0.0, InfiniBand support through the
openibplugin is both deprecated and superseded by the
ucxPML component. The
openibBTL was removed in Open MPI v5.0.0.
openibBTL depended on
libibverbs, the UCX PML depends on the UCX library.
Once installed, Open MPI can be built with UCX support by adding
--with-ucxto the Open MPI configure command. Once Open MPI is configured to use UCX, the runtime will automatically select the
ucxPML if one of the supported networks is detected (e.g., InfiniBand). It’s possible to force using UCX in the
oshruncommand lines by specifying any or all of the following mca parameters:
--mca pml ucxfor MPI point-to-point operations,
--mca spml ucxfor OpenSHMEM support, and
--mca osc ucxfor MPI RMA (one-sided) operations.
usnicBTL is support for Cisco’s usNIC device (“userspace NIC”) on Cisco UCS servers with the Virtualized Interface Card (VIC). Although the usNIC is accessed via the OpenFabrics Libfabric API stack, this BTL is specific to Cisco usNIC devices.
uGNI is a Cray library for communicating over the Gemini and Aries interconnects.
knemsupport is used when the
sm(shared memory) BTL is compiled with knem support (see the
--with-knemconfigure option) and the
knemLinux module is loaded in the running kernel. If the
knemLinux kernel module is not loaded, the
knemsupport is (by default) silently deactivated during Open MPI jobs.
See https://knem.gitlabpages.inria.fr/ for details on Knem.
Linux Cross-Memory Attach (CMA) or XPMEM is used by the
smshared memory BTL when the CMA/XPMEM libraries are installed, respectively. Linux CMA and XPMEM are similar (but different) mechanisms for Open MPI to utilize single-copy semantics for shared memory.
The OFI MTL does not support sending messages larger than the active Libfabric provider’s
max_msg_size. If you receive an error message about sending too large of a message when using the OFI MTL, please reach out to your networking vendor to ask them to support a larger
max_msg_sizefor tagged messages.