Hi Sage To be sure to have the good understanding : if I reached the max number of PG per OSD with for example 4 pools, and I have to create 2 new pools without adding OSD, I need to migrate old pools to less PGs pool, right ? Thanks
Sent from my iPhone > On 23 déc. 2014, at 15:39, Sage Weil <[email protected]> wrote: > >> On Tue, 23 Dec 2014, Ren? Gallati wrote: >> Hello, >> >> so I upgraded my cluster from 89 to 90 and now I get: >> >> ~# ceph health >> HEALTH_WARN too many PGs per OSD (864 > max 300) >> >> That is a new one. I had too few but never too many. Is this a problem that >> needs attention, or ignorable? Or is there even a command now to shrink PGs? > > It's a new warning. > > You can't reduce the PG count without creating new (smaller) pools > and migrating data. You can ignore the message, though, and make it go > away by adjusting the 'mon pg warn max per osd' (defaults to 300). Having > too many PGs increases the memory utilization and can slow things down > when adapting to a failure, but certainly isn't fatal. > >> The message did not appear before, I currently have 32 OSDs over 8 hosts and >> 9 >> pools, each with 1024 PG as was the recommended number according to the OSD * >> 100 / replica formula, then round to next power of 2. The cluster has been >> increased by 4 OSDs, 8th host only days before. That is to say, it was at 28 >> OSD / 7 hosts / 9 pools but after extending it with another host, ceph 89 did >> not complain. >> >> Using the formula again I'd actually need to go to 2048PGs in pools but ceph >> is telling me to reduce the PG count now? > > The guidance in the docs is (was?) a bit confusing. You need to take the > *total* number of PGs and see how many of those per OSD there are, > not create as many equally-sized pools as you want. There have been > several attempts to clarify the language to avoid this misunderstanding > (you're definitely not the first). If it's still unclear, suggestions > welcome! > > sage > > _______________________________________________ > ceph-users mailing list > [email protected] > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com _______________________________________________ ceph-users mailing list [email protected] http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
