Hi everyone, this morning I noticed that all email services were in block, namely Sogo and Rouncube emal. The login was successful but the message list did not appear. I assume that the imap daemon has gone into error. I checked out of Sogo here is an excerpt:
Oct 04 08:27:29 sogod [9428]: <0x0x56099f92ac90 [LDAPSource]> <NSException: 0x56099ff5e940> NAME: LDAPException REASON: operation bind failed: Invalid credentials (0x31) INFO: {“error_code” = 49;
Apache:
[Mon Oct 04 09: 07: 54.123289 2021] [lbmethod_heartbeat: notice] [pid 1220] AH02282: No slotmem from mod_heartmonitor
yes I looked and exposed the logs available. It appears that dovecot could no longer obtain system user credentials. I rebooted and everything came back online. However I would like to understand what went wrong. My server has been virtualized on proxmox for at least 4 years.
Due to error related to user individuation, maybe a full server restart could be avoided with only the restart of NSDC, which is userbase and the LDAP server contacted by any other user-related module.
Moreover, a little check on CPU consumption state could tell you if the container was busting is hat off or was not answering at all…
Hello and thanks, the server has 4 cpu used little, the problem in my opinion is that the disk I / O are high being a VM with ZFS and a mechanical disk unfortunately the latency is sometimes high. I tried to configure the cache on the proxmox parameters but I didn’t get great results, considering that there is a risk of losing data. Since the machine is replicating on 3 nodes, I checked if this could lead to strong slowdowns, but none of this. Thanks for showing me how to act on nsdc, I always check sssd, but as you say nscd refers to all the authentications of the services so we deduce that it is more global as a daemon.
Without replacing the mechanical disk it is possible to improve the performance of the VM with cache or other