Hello,
I noticed that during the up2date when the primary node (default master) is in up2date state and the second node is taking over no client traffic ist routed.
Logging into console on node2 I can ping outside and inside - but no traffic from LAN to WAN is routed.
When primary node is back online and syncing is finished everything works again as expected.
Any idea where I could start investigating?
The only difference I noticed so far during this failing state was:
The "route" command on active node2 does not resolve names. Although pinging external servers via DNS-name is possible and pinging the internal DNS via IP, too.
When node1 is back as master a "route" command does resolve DNS names.
In both cases in /etc/resolf.conf is the only entry 127.0.0.1
This is not as expected but should not disrupt traffic to explicit IP-addresses.
Config:
Software UTM v. 9.603-1
1:1 identical Lenovo Servers for node 1 and 2
"Keep node(s) reserved during Up2Date" is disabled
This thread was automatically locked due to age.