hello,
I have some strange results with 100% Write 100% Random test on windows and Linux system on RAW disk
Result are 8 time lower than IOmeter result or crystaldiskmark
(My other read random/seqential tests seems good)
Currently I'm testing some SSD disk (maybe not the same behavior with SATA disk ? i need to test)
If someone have some tips to fix or improve my tests pattern or diag the issue, thanks !
here my Linux tests params :
*SD: Storage Definitions
sd=sd2,lun=/dev/sdb,threads=2,openflags=o_direct
sd=sd4,lun=/dev/sdb,threads=4,openflags=o_direct
sd=sd8,lun=/dev/sdb,threads=8,openflags=o_direct
sd=sd16,lun=/dev/sdb,threads=16,openflags=o_direct
sd=sd32,lun=/dev/sdb,threads=32,openflags=o_direct
sd=sd64,lun=/dev/sdb,threads=64,openflags=o_direct
sd=sd128,lun=/dev/sdb,threads=128,openflags=o_direct
*WD: Workload Definitions
wd=WD_256k_FULLwriteRND_2,sd=sd2,rdpct=0,seekpct=100,xfersize=256k
wd=WD_256k_FULLwriteRND_4,sd=sd4,rdpct=0,seekpct=100,xfersize=256k
wd=WD_256k_FULLwriteRND_8,sd=sd8,rdpct=0,seekpct=100,xfersize=256k
wd=WD_256k_FULLwriteRND_16,sd=sd16,rdpct=0,seekpct=100,xfersize=256k
wd=WD_256k_FULLwriteRND_32,sd=sd32,rdpct=0,seekpct=100,xfersize=256k
wd=WD_256k_FULLwriteRND_64,sd=sd64,rdpct=0,seekpct=100,xfersize=256k
wd=WD_256k_FULLwriteRND_128,sd=sd128,rdpct=0,seekpct=100,xfersize=256k
*RD: Run Definitions
rd=run_256k_FULLwriteRND_2,wd=WD_256k_FULLwriteRND_2,elapsed=300,iorate=max,interval=20,pause=5
rd=run_256k_FULLwriteRND_4,wd=WD_256k_FULLwriteRND_4,elapsed=300,iorate=max,interval=20,pause=5
rd=run_256k_FULLwriteRND_8,wd=WD_256k_FULLwriteRND_8,elapsed=300,iorate=max,interval=20,pause=5
rd=run_256k_FULLwriteRND_16,wd=WD_256k_FULLwriteRND_16,elapsed=300,iorate=max,interval=20,pause=5
rd=run_256k_FULLwriteRND_32,wd=WD_256k_FULLwriteRND_32,elapsed=300,iorate=max,interval=20,pause=5
rd=run_256k_FULLwriteRND_64,wd=WD_256k_FULLwriteRND_64,elapsed=300,iorate=max,interval=20,pause=5
rd=run_256k_FULLwriteRND_128,wd=WD_256k_FULLwriteRND_128,elapsed=300,iorate=max,interval=20,pause=5