Forums

Home / Forums

You need to log in to create posts and topics. Login · Register

1 client failing to respond to cache pressure.

how to fix it?

try increasing mds memory, example

ceph config set mds mds_cache_memory_limit 4GB

How to know my current mds cache limit

ceph config get mds mds_cache_memory_limit

currently i have a new issue""lease expired failed on NFSv4 server ip with error 121"" what does this means how to overcome it?

Hi,

I have very wired situation. It begins just after update cluster to 3.0.0, and still present in 3.2.0

My HEALTH_WARN            1 clients failing to respond to cache pressure
exists almost all the time.

I tried to set cache from 4 GB to 6 GB and 8 GB but it does not help.

Result below says, that cache is used 0,15 GB and 0,08 GB.

ceph tell mds.* cache status
2023-07-28T12:15:19.103+0200 7fd4c17fa700  0 client.91857905 ms_handle_reset on v2:172.30.0.42:6800/623421266
2023-07-28T12:15:19.135+0200 7fd4c17fa700  0 client.91823569 ms_handle_reset on v2:172.30.0.42:6800/623421266
Error ENOSYS:
2023-07-28T12:15:19.139+0200 7fd4c17fa700  0 client.91823575 ms_handle_reset on v2:172.30.0.41:6800/1186658423
2023-07-28T12:15:19.167+0200 7fd4c17fa700  0 client.91823581 ms_handle_reset on v2:172.30.0.41:6800/1186658423
mds.ceph01: {
"pool": {
"items": 5962916,
"bytes": 151031037
}
}
2023-07-28T12:15:19.175+0200 7fd4c17fa700  0 client.91823587 ms_handle_reset on v2:172.30.0.43:6896/2544779998
2023-07-28T12:15:19.203+0200 7fd4c17fa700  0 client.91823593 ms_handle_reset on v2:172.30.0.43:6896/2544779998
mds.ceph03: {
"pool": {
"items": 2904338,
"bytes": 89250706
}
}

What other thing can resolve this issue?

This problem has gone after upgrade to 4.x (for me)