[SOLVED] Ping OK but cluster loosing connectivity

[AxelTwin]

Hi there,
Yesterday I added a third node to my cluster and since then the quorum keeps going up and down randomly.
both nodes "hyperviser2" and "hyperviser" are on the same site same network, and the new one "hyperbackup1" is on a remote site, connected through an ipsec tunnel on a different network. I tried to monitor the connectivity and ping is ok.

Any idea where to look as it doesn't seem to be a routing issue



even between the 2 hypervisers directly connected ("hyperviser2" and "hyperviser") I get some issues.

root@hyperbackup1 ~ # traceroute 192.168.1.110
traceroute to 192.168.1.110 (192.168.1.110), 30 hops max, 60 byte packets
 1  192.168.0.2 (192.168.0.2)  0.157 ms  0.147 ms  0.145 ms
 2  * * *
 3  192.168.1.110 (192.168.1.110)  32.125 ms  32.123 ms  32.121 ms
 
 
 traceroute to 192.168.0.1 (192.168.0.1), 30 hops max, 60 byte packets
 1  pfsense.eec31.local (192.168.1.1)  0.179 ms  0.179 ms  0.152 ms
 2  * * *
 3  hyperbackup1.eec31.local (192.168.0.1)  32.118 ms  32.111 ms  32.103 ms

 

[cheiss]

Hi,

clustering requires a low-latency network, at best with under 5ms latency. Thus, setting up a non-LAN cluster is discouraged or will not work all that well. See also the Cluster Network documentation for more details.

According to your ping/traceroute logs, there is >30ms latency between between both sites - that is way too much for corosync.

 

[AxelTwin]

Ok, got it
Thanks for the hint

 

[cheiss]

Please just mark the thread as solved by editing the first post - there should be a dropdown near
the title field. This helps other people with the same problem to more easily find this thread in
the future!