Out of curiosity do you have compression enabled? FWIW Deepika is has been working on splitting the mempool assignments into much better categories for better tracking. I suspect we are going to find a bug where something isn't being cleaned up properly in buffer_anon. Adam's been taking up that effort. Here's Deepika's PR for better mempool tracking: https://github.com/ceph/ceph/pull/34537 On 5/21/20 5:48 AM, aoanla@gmail.com wrote:
So, to jump into this thread - we seem to see the same problem as Harald on our cluster here in Glasgow, except our "worst case" OSDs are much worse than his [we get up to ~tens of GB in buffer_anon].
Activity is a mix of reads and writes against a single EC (8+2) encoded pool, with 8MB objects. _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io