2015-10-10 02:29 PM
WHen you halt a node in a two node cluster , should there be any downtime if the cluster LIF is on the node that goes down.
I would assume a second or two while the LIF migrates to the other node , or am i wrong ?
Reason for asking have a two node cluster and am seeing the interface go down , if i migrate the lif before halting node I dont see downtime
Lifs from both nodes are in the same domain , and all looks ok
is there any thing I could have possibly missed or I should check
Cheers for any info
2015-10-11 10:18 AM
If you constant ping the cluster address and the node that the cluster LIF sits on fails , ie memory dump will you lose connectivity to the cluster address while the LIF moves to the other node , and If so, how long
In this type of senario i am seening the LIF not failover , but if i do a LIF migrate it migrates ok to the other node , I was under the impression that the cluster address should at most lose a couple of pings while it failsover or am i wrong
cheers for any info
2015-10-11 09:00 PM
2015-10-20 03:04 PM
so found the issue , networking
the managment ports where plugged into juniper ports and the equivilent portfast was not set , so i was seeing an outage of the cluster management address for at least a min , now switch configured , during a failover i see 1 ping drop