empty ALL rrd graph "summary" in web - while remote NAS unavailable

[Nesterri]

sorry for my English (


I have a problem:
to proxmox-node attached 3 storage (nfs for backup), while they are on a remote NAS, if there is no connection to this NAS - in the web-interface proxmox, in ALL tabs in ALL VM "summary" (rrd graph) empty value lines.

syslog writes:
"Oct 27 09:38:03 hostvirt pveproxy [558,633]: proxy detected vanished client connection
Oct 27 09:38:30 hostvirt kernel: ct0 nfs: server 192.168.6.249 not responding, still trying."


i.e. stop features beautiful graphics for all the VM, in the case of loss of connection with the NAS, although VM continue to work steadily.


maybe somehow fix it? or turn off rrd-graph all storage NAS?

 

[Nesterri]

up (

 

[vkhera]

That is more or less the best you can do when you have no data for that time.

However, I will note that the monitoring pings the NFS server once every 10 seconds from each and every node when the resolution of the charts is much less than that. I would prefer that the monitor be less chatty, but that is really low on my list of things that need improvement.

 

[wolfgang]

Hi,

this is a known issue and we have at the moment no solution for it.
this is a very complicated to solve, but we are on it.

 

[wahmed]

Just wanted to confirm that I am also seeing this issue. But oddly graph is not empty for all nodes in the cluster. In my case there were only 2 nodes that did not show rrd graphs for several days. As soon as i deleted the faulty NFS share graphs showed for these 2 nodes within minutes.