From d53f6ebfa308533944bfca5663cd7111ded1908d Mon Sep 17 00:00:00 2001 From: pearce8 Date: Mon, 4 May 2026 15:22:03 -0500 Subject: [PATCH] Initial docs for IOR --- docs/80_ior/ior.rst | 60 +++++++++++++++++++++++++++++++++++---------- 1 file changed, 47 insertions(+), 13 deletions(-) diff --git a/docs/80_ior/ior.rst b/docs/80_ior/ior.rst index a0907ad..54dca4a 100644 --- a/docs/80_ior/ior.rst +++ b/docs/80_ior/ior.rst @@ -2,35 +2,73 @@ IOR *** -https://github.com/hpc/ior - - Purpose ======= +IOR is used for testing performance of parallel file systems using various interfaces and access patterns at the POSIX and MPI-IO level. Characteristics =============== -Problems --------- +IOR is available in the benchmarks repository. + +* Github: `IOR Public `_ -Figure of Merit ---------------- +The github repo also contains mdtest. +Problem +------- + +IOR measures parallel I/O performance at the POSIX and MPI-IO levels. +It writes and reads files, one per rank or shared between all ranks, on a parallel file system. +It should be run in lustre space. Source code modifications ========================= Please see :ref:`GlobalRunRules` for general guidance on allowed modifications. + Building ======== +MPI, MPI-IO, and OpenMP are required in order to build and run the code. The +source code used for this benchmark is derived from IOR 3.0.1 and it is +included here. + +Ensure that the MPI compiler wrappers (e.g., `mpicc`) are in `$PATH`. Then create a build directory and an (optional) install directory. + +.. code-block:: bash + + /microbenchmarks/ior/configure --prefix= + make + #make install +.. + +This will build both IOR with the POSIX and MPI-IO interfaces and create the +IOR executable at `src/ior`. Running ======= +The ior tests can be run using the following command: + +.. code-block:: bash + + srun ... + +.. + +Where `load_type` is `load1` for sequential loads and `load2` for random loads, `io_type` is `posix` or `mpiio`, and `access_type` is `filepertask` and `sharedfile` for per task and shared accesses respectively. +There are six input decks in the `inputs.xroads` directory; each should be run on a single node and across the full system in parallel. + +"*Note: Benchmark values for random loads are not presented here.*" + +Input +----- + +Running IOR does not require using the input files. All arguments can be given on the command line. + Validation ========== @@ -44,12 +82,8 @@ Memory Usage ============ -Strong Scaling on El Capitan -============================ - - -Weak Scaling on El Capitan -========================== +Scaling on El Capitan +===================== References