[ceph-users] client failing to respond to cache pressure

[Zhenshi Zhou]

Hi

I used to get warning message claims client faling to respond
to cache pressure. After I switch rockdb and wal data to ssd, the
message seems disappeared.

However it shows again yesterday and the message looks a little
different: *MDS_CLIENT_RECALL_MANY: 1 MDSs have many *
*clients failing to respond to cache pressure.*

I added some fuse client and it has 48 clients mounting the cephfs
at present. The cluster has only one MDS active and one standby.

How can I get the cluster healthy again? I'm happy to provide any
information of the cluster.

Thanks

[Zhenshi Zhou]

Hi

The cluster recovers healthy after I set "mds_cache_memory_limit" from 4G
to 8G.

Post by Zhenshi Zhou
Hi
I used to get warning message claims client faling to respond
to cache pressure. After I switch rockdb and wal data to ssd, the
message seems disappeared.
However it shows again yesterday and the message looks a little
different: *MDS_CLIENT_RECALL_MANY: 1 MDSs have many *
*clients failing to respond to cache pressure.*
I added some fuse client and it has 48 clients mounting the cephfs
at present. The cluster has only one MDS active and one standby.
How can I get the cluster healthy again? I'm happy to provide any
information of the cluster.
Thanks