Mark Nelson
Mark Nelson
> > @jianwei1216 ah! ok. I don't have an HDD test bed handy at the moment. I'm really surprised that you were seeing ~100% CPU time in the msgr threads...
@lee-j-sanders Thanks a ton for testing this! @jianwei1216 Sorry it took so long to test this. Once @lee-j-sanders has confirmed, I think we can move forward and get this into...
retest this please
@aclamk Very nice! Digging those 1.7 results especially. They are nearly as good as ref in write IOPS and significantly better in all other metrics.
@pereman2 I was very excited about your PR, and did some librbd fio testing on mako over the weekend. These are relatively fast NVMe drives so I'm not sure how...
> @markhpc I'm curious. How many times did you run the 4k randwrite benchmark? Did you pre fill the cluster to simulated some real data or was it ran in...
> > > @markhpc I'm curious. How many times did you run the 4k randwrite benchmark? Did you pre fill the cluster to simulated some real data or was it...
@pereman2 can you provide the specs for the HW used for that testing please? Thanks! Edit: It's unusual to see bluestore performing the same for read/write IOPS. For instance on...
> @markhpc db devices: > > ``` > osd 0 -> nvme1n1 INTEL SSDPE2KX080T8 7.3T > osd 1 -> nvme2n1 INTEL SSDPE2KX080T8 7.3T > osd 2 -> nvme3n1 INTEL SSDPE2KX080T8...
> I'm not sure "data_layout_reformatting" is very clear to the users what it is doing. > > Maybe instead use "optimize_during_scrub" with options "defrag", "recompress", "all", ... Or perhaps better:...