No other wdqs servers have that same behaviour, only wdqs1005. And this means that we are running out of disk space.
Some investigation needs to be done and then probably copy data over from another node.
No other wdqs servers have that same behaviour, only wdqs1005. And this means that we are running out of disk space.
Some investigation needs to be done and then probably copy data over from another node.
Mentioned in SAL (#wikimedia-operations) [2019-08-05T19:51:12Z] <gehel> depool wdqs1005 - T229876
The journal dump says: There are 755560 commit points.. Probably not a normal situation.
@Gehel I think for now we need to reload the DB from other server and repool it. Let's keep an eye on this problem and see if it ever happens again.