I am having difficulty when adding a new disk to a vm where the storage selector dropdown is timing out. I assume this is from I/O degradation so the dropdown can't get the list of available NFS or iSCI mounts.
But I can't ascertain the cause.
The NAS is connected via quad gigabit bond and the current network utilization on that interface is < 100Mbps.
This is on pve-manager/4.3-1/e7cdc165 (running kernel: 4.4.19-1-pve) using a two-server cluster. If I move the VM to the other server, it exhibits the same behavior.
My "guess" is this is some performance issue with NFS, but I don't know where to look to find out.
Also, not sure if this is related, but I happened to leave an ssh session open last night and had this on one of the servers this morning:
Broadcast message from systemd-journald@ucs (Wed 2016-10-26 00:17:23 EDT):
iscsid[2042]:
Message from syslogd@ucs at Oct 26 00:17:23 ...
iscsid:
Broadcast message from systemd-journald@ucs (Wed 2016-10-26 05:46:18 EDT):
iscsid[2042]:
Message from syslogd@ucs at Oct 26 05:46:18 ...
iscsid:
Quite an odd message (no message at all, just "iscid:").
But I can't ascertain the cause.
The NAS is connected via quad gigabit bond and the current network utilization on that interface is < 100Mbps.
This is on pve-manager/4.3-1/e7cdc165 (running kernel: 4.4.19-1-pve) using a two-server cluster. If I move the VM to the other server, it exhibits the same behavior.
My "guess" is this is some performance issue with NFS, but I don't know where to look to find out.
Also, not sure if this is related, but I happened to leave an ssh session open last night and had this on one of the servers this morning:
Broadcast message from systemd-journald@ucs (Wed 2016-10-26 00:17:23 EDT):
iscsid[2042]:
Message from syslogd@ucs at Oct 26 00:17:23 ...
iscsid:
Broadcast message from systemd-journald@ucs (Wed 2016-10-26 05:46:18 EDT):
iscsid[2042]:
Message from syslogd@ucs at Oct 26 05:46:18 ...
iscsid:
Quite an odd message (no message at all, just "iscid:").