1. has been discussed multiple times in the last weeks, for example here:
https://lists.ceph.io/hyperkitty/list/ceph-users@ceph.io/thread/S5XBS63NFIBJHN44RQ4NES7F6GVIMH4T/
2. deviation means in both directions. The problem is, it doesn't
distinguish between OSD sizes. Larger OSDs will naturally get more PGs
assigned, but the current check doesn't take the sizes into account. A
couple of months ago, we figured out a way to get the distinction into
the existing mgr code without modfying too much. We're waiting for the
customer to either complain or confirm that it works as intended. :-)
Zitat von Devender Singh <deven...@netskrt.io>:
Hello all
Seeking some help to solve below…
1. # ceph health detail
HEALTH_WARN 2 OSD(s) experiencing slow operations in BlueStore
[WRN] BLUESTORE_SLOW_OP_ALERT: 2 OSD(s) experiencing slow operations
in BlueStore
osd.10 observed slow operation indications in BlueStore
osd.21 observed slow operation indications in BlueStore
Having some other OSD’s with same size with more PG’s but only
complaining on below…
# ceph osd df |sort +17|egrep -i "^10|^21"
21 ssd 6.98630 1.00000 7.0 TiB 2.7 TiB 2.6 TiB 17 MiB
27 GiB 4.3 TiB 38.22 0.85 103 up
10 ssd 6.98630 1.00000 7.0 TiB 4.0 TiB 4.0 TiB 21 MiB
35 GiB 2.9 TiB 57.97 1.28 123 up
2. What does deviate means towards? holding higher number of pgs or Lower?
OSD osd.6 on host1 deviates by more than 30% from average PG count.
OSD osd.5 onhost2 deviates by more than 30% from average PG count.
# ceph osd df |sort +17|egrep -i "^ 5|^ 6”
5 ssd 6.98630 1.00000 7.0 TiB 2.8 TiB 2.8 TiB 11 MiB 32 GiB 4.2
TiB 39.91 0.88 97 up
6 ssd 6.98630 1.00000 7.0 TiB 3.2 TiB 3.1 TiB 19 MiB 48 GiB 3.8
TiB 45.65 1.01 103 up
Regards
Dev
_______________________________________________
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io
_______________________________________________
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io
