[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-30 Thread mahnoosh shahidi
Hi Mark, Thanks for your response. I did manual compaction on all osds using ceph-kvstore-tool. It reduced the number of slow ops but It didn't solve the problem completely. On Mon, Jul 26, 2021 at 8:06 PM Mark Nelson wrote: > Yeah, I suspect that regular manual compaction might be the necessary

[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-26 Thread Mark Nelson
Yeah, I suspect that regular manual compaction might be the necessary work around here if tombstones are slowing down iterator performance.  If it is related to tombstones, it would be similar to what we saw when we tried to use deleterange and saw similar performance issues. I'm a little at a

[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-26 Thread Igor Fedotov
Hi Mahnoosh! Unfortunately I'm not an expert in RGW hence nothing to recommend from this side. Apparently your issues are caused by bulk data removal - it appears that RocksDB can hardly sustain such things and its performance degrades. We've seen that plenty of times before. So far there

[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-26 Thread mahnoosh shahidi
Thanks for your help. Our hdd osd have separate nvme disks for DB use. On Mon, Jul 26, 2021 at 3:49 PM Igor Fedotov wrote: > Unfortunately I'm not an expert in RGW hence nothing to recommend from > that side. > > Apparently your issues are caused by bulk data removal - it appears that > RocksDB

[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-26 Thread Igor Fedotov
Unfortunately I'm not an expert in RGW hence nothing to recommend from that side. Apparently your issues are caused by bulk data removal - it appears that RocksDB can hardly sustain such things and its performance degrades. We've seen that plenty of times before. So far there are two known w

[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-25 Thread mahnoosh shahidi
Hi Igor, Thanks for your response.This problem happens on my osds with hdd disks. I set the bluefs_buffered_io to true just for these osds but it caused my bucket index disks (which are ssd) to produce slow ops. I also tried to set bluefs_buffered_io to true in bucket index osds but they filled the

[ceph-users] Re: octopus garbage collector makes slow ops

2021-07-21 Thread Igor Fedotov
Hi Mahnoosh, you might want to set bluefs_buffered_io to true for every OSD. It looks it's false by default in v15.2.12 Thanks, Igor On 7/18/2021 11:19 PM, mahnoosh shahidi wrote: We have a ceph cluster with 408 osds, 3 mons and 3 rgws. We updated our cluster from nautilus 14.2.14 to octopu