Ceph » bluestore

Igor Fedotov wrote:

https://github.com/ceph/ceph/pull/30144

merged

Alexander Patrakov wrote:

Nathan Cutler wrote:

This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/30144
merge commit 95bbbc09d6caab8d6063373dffb0bdd73ecdae43 (v14.2.4-782-g95bbbc09d6)

We have a customer who hit this assert on 14.2.8, where it is supposed to be already fixed. Last 50000 lines from the log are attached. Will also ask for ceph.conf.

Asked, nothing suspicious there except the fact that they still use separate roots instead of device classes:

[global]
cluster network = 10.15.0.0/24
fsid = 8700e000-d2ac-4393-8380-1cf4779166b5
mon host = [v2:10.15.0.201:3300,v1:10.15.0.201:6789],[v2:10.15.0.202:3300,v1:10.15.0.202:6789],[v2:10.15.0.203:3300,v1:10.15.0.203:6789],[v2:10.15.0.204:3300,v1:10.15.0.204:6789],[v2:10.15.0.205:3300,v1:10.15.0.205:6789]
mon initial members = ceph01,ceph02,ceph03,ceph04,ceph05
osd pool default crush rule = -1
public network = 10.15.0.0/24

[osd]
osd memory target = 11811762995
osd crush update on start = false
osd scrub begin hour = 23
osd scrub end hour = 3
osd scrub load threshold = 0.3

@Alexander - it might make sense to open a new bug in the Bluestore project for that, since this one is closed.

Nathan Cutler wrote:

@Alexander - it might make sense to open a new bug in the Bluestore project for that, since this one is closed.

Just created https://tracker.ceph.com/issues/48216 for that