Benchmark programs
- IOR (http://sourceforge.net/projects/ior-sio) has a 'driver' for pnetcdf
- The FLASH I/O kernel (http://www-unix.mcs.anl.gov/pio-benchmark/code/flash-argonne-20060913.tar.gz) will measure several write-only workloads
Notes and suggestions
Be mindful of benchmarks which write large contiguous chunks of data. While interesting, it might not be what a scientific application using pnetcdf would actually do.
Older versions of IOR triggered a performance bug on some Lustre file systems. Sometimes slow write performance on Lustre can be fixed by disabling the "data sieving" optimization. See HintsForPnetcdf for more information.
![(please configure the [header_logo] section in trac.ini)](/projects/parallel-netcdf/chrome/common/trac_banner.png)