White Papers

Dell Storage for HPC with Intel Enterprise Edition for Lustre sofware
Appendix A: Benchmark Command Reference
This section describes the commands used to benchmark the Dell Storage for HPC with Intel EE Lustre
solution.
IOzone
IOzone Sequential Writes –
iozone -i 0 -c -e -w -r 1024K -I -s $Size -t $Thread -+n -+m /root/list.$Thread
IOzone Sequential Reads -
iozone -i 1 -c -e -w -r 1024K -I -s $Size -t $Thread -+n -+m /root/list.$Thread
IOzone IOPS Random Reads / Writes –
iozone -i 2 -w -c -O -I -r 4K -s $Size -t $Thread -+n -+m /root/list.$Thread
IOzone Command Line
Arguments
Description
-i 0
Write test
-i 1
Read test
-i 2
Random IOPS test
-+n
No retest
-c
Includes close in the timing calculations
-e
Includes flush in the timing calculations
-r
Records size
-s
File size
-+m
Location of clients to run IOzone on when in clustered mode
-I
Use O_Direct
-w
Does not unlink (delete) temporary file
-+n
No retests selected
-O
Return results in OPS
The O_Direct command line parameter (“-I”) allows us to bypass the cache on the compute nodes
where the IOzone threads are running.
IOR
IOR Writes –
mpirun -np $Threads -rr --machinefile /root/list.$Threads /cm/share/IOR -a POSIX -v -i $rep -d 3
-e -k -o /mnt/boulder/perf/mytestfile -w -s 1 -t 4m -b $SizePerThread
IOR Reads –
mpirun -np $Threads -rr --machinefile /root/list.$Threads /cm/share/IOR -a POSIX -v -i $rep -d 3
-e -k -o /mnt/boulder/perf/mytestfile -r -s 1 -t 4m -b $SizePerThread
IOR
Command Line Arguments
Description
-a S
api -- API for I/O [POSIX|MPIIO|HDF5|NCMPI]
-v
verbose -- output information (repeating flag increases level)
-I N
repetitions -- number of repetitions of test
-d N
interTestDelay -- delay between reps in seconds
-e
fsync -- perform fsync upon POSIX write close
-k
keepFile -- don't remove the test file(s) on program exit
-o S
testFile -- full name for test
-w
writeFile -- write file
-r
readFile -- read existing file
-s N
segmentCount -- number of segments
-t N
transferSize -- size of transfer in bytes (e.g.: 8, 4k, 2m, 1g)
-b N
blockSize -- contiguous bytes to write per task (e.g.: 8, 4k, 2m, 1g)