Hello Users, I've also got a cluster upgraded from v18.2.0 to v18.2.1 recently where it appears to be working fine. Not sure if I'm missing something here. ceph> version ceph version 18.2.1 (7fe91d5d5842e04be3b4f514d6dd990c54b29c76) reef (stable) ceph> status cluster: id: 80468512-289d-11ee-a043-314ee4b0dffc health: HEALTH_OK services: mon: 1 daemons, quorum cs418 (age 6d) mgr: cs418.qihtft(active, since 6d), standbys: cs418.wksgkz mds: 1/1 daemons up, 1 standby osd: 25 osds: 25 up (since 6d), 25 in (since 5M) data: volumes: 1/1 healthy pools: 5 pools, 1201 pgs objects: 212.12k objects, 814 GiB usage: 1.6 TiB used, 33 TiB / 35 TiB avail pgs: 1201 active+clean io: client: 39 KiB/s rd, 979 KiB/s wr, 50 op/s rd, 158 op/s wr ceph> osd pool autoscale-status POOL SIZE TARGET SIZE RATE RAW CAPACITY RATIO TARGET RATIO EFFECTIVE RATIO BIAS PG_NUM NEW PG_NUM AUTOSCALE BULK .mgr 14148k 2.0 35909G 0.0000 1.0 1 on False .nfs 13024 2.0 35909G 0.0000 1.0 32 on False cephfs.cloudstack.meta 24290k 2.0 35909G 0.0000 4.0 16 on False cephfs.cloudstack.data 14431M 2.0 35909G 0.0008 1.0 128 off False cloudstack 771.0G 2.0 35909G 0.0429 1.0 1024 on True Thanks, Jayanth On Mon, Dec 25, 2023 at 9:27 PM Jayanth Reddy <jayanthreddy5666@gmail.com> wrote:
Hello Users, I deployed a new cluster with v18.2.1 but noticed that pg_num and pgp_num always remained 1 for the pools with autoscale turned on. Below is the env and the relevant information
ceph> version ceph version 18.2.1 (7fe91d5d5842e04be3b4f514d6dd990c54b29c76) reef (stable) ceph> status cluster: id: 273c8410-a333-11ee-b3c2-9791c3098e2b health: HEALTH_WARN clock skew detected on mon.ec-rgw-s3
services: mon: 3 daemons, quorum ec-rgw-s1,ec-rgw-s2,ec-rgw-s3 (age 48m) mgr: ec-rgw-s1.icpgxx(active, since 69m), standbys: ec-rgw-s2.quzjfv osd: 3 osds: 3 up (since 29m), 3 in (since 49m) rgw: 1 daemon active (1 hosts, 1 zones)
data: pools: 8 pools, 23 pgs objects: 1.85k objects, 6.0 GiB usage: 4.8 GiB used, 595 GiB / 600 GiB avail pgs: 23 active+clean
ceph> osd pool get noautoscale noautoscale is off ceph> osd pool autoscale-status ceph> osd pool autoscale-status ceph> osd pool ls detail pool 1 '.mgr' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 21 flags hashpspool stripe_width 0 pg_num_max 32 pg_num_min 1 application mgr read_balance_score 3.00 pool 2 'default.rgw.buckets.data' erasure profile ec-21 size 3 min_size 2 crush_rule 1 object_hash rjenkins pg_num 16 pgp_num 16 autoscale_mode off last_change 111 lfor 0/0/55 flags hashpspool stripe_width 8192 compression_algorithm lz4 compression_mode force application rgw pool 3 '.rgw.root' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 34 flags hashpspool stripe_width 0 application rgw read_balance_score 3.00 pool 4 'default.rgw.log' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 37 flags hashpspool stripe_width 0 application rgw read_balance_score 3.00 pool 5 'default.rgw.control' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 39 flags hashpspool stripe_width 0 application rgw read_balance_score 3.00 pool 6 'default.rgw.meta' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 41 flags hashpspool stripe_width 0 pg_autoscale_bias 4 application rgw read_balance_score 3.00 pool 7 'default.rgw.buckets.index' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 44 flags hashpspool stripe_width 0 pg_autoscale_bias 4 application rgw read_balance_score 3.00 pool 8 'default.rgw.buckets.non-ec' replicated size 3 min_size 2 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 47 flags hashpspool stripe_width 0 application rgw read_balance_score 3.00
I'd manually changed for pool ID 2.
Is this by any chance due to PR [1]?
[1] https://github.com/ceph/ceph/pull/53658
Thanks, Jayanth