RHEV-M running as a HA Cluster
I am trying to setup RHEV-M to run as a HA cluster. I am following a document I found from this site to setup RHCS HA cluster.
I have 2 hosts running ricci and RHEV-M running on a HA-LVM. I am not going to use this host for any other purpose other than running RHEV-M. Reading through the documentation it is not clear if I need a third host to run the web ui luci. Is this true? Do I really need 3 hosts?
If i run luci on one of the hosts and if for some reason that host is down, do i lose the cluster failover capabilities? I can understand if i will not be able to make any cluster changes while luci is down. Will this impact the function of RHEV-M in any way?
Also my understanding is once the cluster is setup the way i need it, i dont have to login to luci for any reason.. Is this a safe assumption? If so can i shutdown luci once the clusters are configured?
When a failover happens i see messages letting logged in /var/log/messages. Is there a way to send these out as SNMP traps?
thanks
Ramesh
Responses
Ramesh,
Think of Luci as a tool to generate /etc/cluster/cluster.conf.
Once that has been completed, you can disable Luci, and operate the cluster using the clustat and clusvcadm commands.
There are lots of ways to turn log messages into SNMP traps. Take a look at the man page for rsyslog.conf to see ways of processing log messages. Or you could just use scripts that invoke the snmptrap command as part of a service definition.