Hi, In our cluster we only the data pool is on ec 4:2, the others are on replica 3.
--- RAW STORAGE --- CLASS SIZE AVAIL USED RAW USED %RAW USED nvme 10 TiB 10 TiB 121 MiB 408 GiB 3.80 ssd 524 TiB 440 TiB 84 TiB 84 TiB 16.01 TOTAL 534 TiB 450 TiB 84 TiB 84 TiB 15.77 --- POOLS --- POOL ID PGS STORED OBJECTS USED %USED MAX AVAIL device_health_metrics 1 1 11 MiB 55 32 MiB 0 124 TiB .rgw.root 2 32 1.0 MiB 178 4.5 MiB 0 124 TiB hkg.rgw.log 22 32 22 GiB 56.35k 65 GiB 0.02 124 TiB hkg.rgw.control 23 32 2.6 KiB 8 7.8 KiB 0 124 TiB hkg.rgw.meta 24 8 322 KiB 1.08k 13 MiB 0 3.2 TiB hkg.rgw.otp 25 32 0 B 0 0 B 0 124 TiB hkg.rgw.buckets.index 26 128 134 GiB 58.57k 403 GiB 3.99 3.2 TiB hkg.rgw.buckets.non-ec 27 32 18 MiB 189.31k 2.2 GiB 0 124 TiB hkg.rgw.buckets.data 28 32 42 TiB 300.06M 63 TiB 14.50 247 TiB As you can see the data pool pg number is 32. If I calculate like max pg for the data pool with 36 OSD should be 512 or let's say 256 because other pools have replica 3, 524/256=2PG/TiB so if stored 84TiB it should have already warned to increase the pg in that pool isn't it? Or EC autoscale doesn't work? Thank you _______________________________________________ ceph-users mailing list -- [email protected] To unsubscribe send an email to [email protected]
