Hi,
I'm running Jetstress on a Vmware system with 10 x 10K RPM SAS drives. The configuration is RAID 10, VMDK eager zeroed, 64K NTFS allocation unit size. Using IO Meter my test results (depending on # of oustanding IOs and read/write access specification) hit anywhere from 800 to 1,500 IOPs.
However when I do the same with JetStress, following the field guide using 2 worker threads I'm getting 241 IOPs, with the test failing due to:
| "Fail | The process has average log write latencies higher than 10.000 msec." |
I'm assuming the low IOPs is due to the number of threads (although I did use the Jetstress field report's guidelines).
I'm also assuming the log write latencies could be caused by the RAID controller (standard issue controller - rather than an upgrade).
I'd be interested in hearing from anyone who has had similar experiences doing comparison tests with IO Meter and JetStress.
Thanks
IT Support/Everything