[Linux-cluster] Cluster and Network outage

Paras pradhan pradhanparas at gmail.com
Thu Aug 20 22:23:50 UTC 2009


Yesterday for around 14 minutes we have a network outage. Today
morning I saw that my redhat Linux cluster was stopped and had to
start the cluster from Conga. It was fine afterwards. I have a qdiskd
in which I  am using heuristics as well. When there was an outage I
think it was not able to ping my router (which I added to heuristics).
Does this bring my whole cluster down?

In Log I don't see anything interesting..

It says heuristics ip --- DOWN

and this
Aug 19 22:52:47 cvtst1 ccsd[3310]: Unable to connect to cluster
infrastructure after 210 seconds.

We usually have network outages. (2 or 3 in a month) How to get rid of this?

And does the network outage brings the cluster to stopped stage?

Thanks!
Paras.




More information about the Linux-cluster mailing list