[ClusterLabs] ethmonitor RA agent error. How can I fix this? (RHEL)

[Ken Gaillot]

On Tue, 2018-05-22 at 09:15 +0800, Confidential Company wrote:
> I have two Virtual machines with two network interfaces. 
> 
> See configuration below:
> 
> *eth0 - service network
> *eth1 - heartbeat network
> 
> *vi /etc/hosts - RhelA(ip of eth1) / RhelB(ip of eth1) 
> *service firewalld stop
> 
> *pcs resource create VirtualIP ocf:heartbeat:IPaddr2 ip=(virtual ip)
> cidr_netmas=32 op monitor interval =30s
> 
> *pcs resource create eth1-monitor ethmonitor interface=eth1 --clone
> 
> *pcs constraint location VirtualIP rule score=-INFINITY ethmonitor-
> ens192 ne 1
> 
> +++++++++++
> 
> I tried to ifdown eth0(service network). The result is:

FYI, corosync 2 cannot recover from ifdown + ifup (which does not
really simulate an interface failure). For testing, it's closer to a
real failure to pull the physical cable, or use the firewall to block
all traffic on the interface.

> 1. VirtualIP resource switched to Node2
> 2. Got an error from pcs status, Error "unable to find nic...."
> 3. Even after a successful failover, error still exist

Do you mean the failure listing in status output? Or maybe log messages
when the policy engine runs? In both cases, that's expected -- it's a
record of a previous failure, not a report of a new failure.

You can clear status output and most policy engine messages by cleaning
up the resource (preferably after investigating and fixing whatever
caused the error, otherwise it's likely to cause further issues).

> Since it automatically switced to Node2, my goal is to failover again
> to Node1. This is what I did:
> 
> 1. Enable eth0 of node1, wait for 15 seconds.
> 2. Disable eth0 of node2
> 2. VirtualIP resource got stopped
> 3. Even after enabling eth0 of node1, error from previous procedure
> still exist.
> 4. Got an additional error, I have two errors now
> 5. VirtualIP resource doesn't start
> 
> 
> Regards,
> 
> imnotarobot
-- 
Ken Gaillot <kgaillot at redhat.com>