topic Re: Cluster alert in ExtremeWireless (General)

Cluster alert

Jeff_Roberts — Sat, 08 Jun 2019 02:54:41 GMT

One of my rfs6000 wing controllers is logging this alert every two minutes:

Jun 07 14:39:48 2019: wls-dc07-1 : %CLUSTER-4-STATE_RETAIN_ACTIVE: Member 19.F9.60.CB (load50) retaining Active state. New member 19.F9.5C.43 standby load 8.

Here are cluster members:
wls-dc07-1#sh clu mem
------------------------------------------------------------------------------------------
HOSTNAME MEMBER-ID MAC MASTER OPERATIONAL-STATE LAST-SEEN
------------------------------------------------------------------------------------------
wls-dc07-1 19.F9.60.CB B4-C7-99-F9-60-CB True active self
wls-dc07-2 19.F9.5C.43 B4-C7-99-F9-5C-43 False active 00:01:44 ago
------------------------------------------------------------------------------------------

I don't have that alert selected for anything in the Event Policy. Not sure why it's telling me every two minutes that it's retained its active state.

Here are my logging settings:
Logging module: enabled
Aggregation time: disabled
Console logging: level warnings
Monitor logging: disabled
Buffered logging: level warnings
Syslog logging: level warnings

All my controllers have the same Event policy settings. None of my other controllers are logging this.

Any ideas?

Thanks!
Jeff Roberts

Re: Cluster alert

ckelly — Mon, 10 Jun 2019 23:31:32 GMT

Jeff, for clusters, I'd expect that you would only be seeing this message on the 'active' cluster member on any clusters you have setup.
So, when you say you don't see this message on the other controllers, are these other controllers 'Primary' active cluster members too?

What version of WiNG?

Something that I'm curious about is the out "New member 19.F9.5C.43 standby load 8" Can you confirm that the MiNT ID 19.F9.5C.43 is the correct MiNT ID for the cluster member? The "New" part of this message is what's throwing me.

Can you run these and post the results for this particular cluster?

show cluster configuration (on both cluster members)
show cluster status (on both cluster members)
show cluster history (on the controller generating these messages)

Re: Cluster alert

Jeff_Roberts — Tue, 11 Jun 2019 04:50:00 GMT

Thanks Chris,
I had an issue with excessive cpu load on several APs connected to the master controller. I attempted to clear the crash-info on the APs to no avail. I reloaded the controller and the crash-info cleared on the APs.
Since the reload, the alert I was asking about : "%CLUSTER-4-STATE_RETAIN_ACTIVE: Member 19.F9.60.CB (load50) retaining Active state. New member 19.F9.5C.43 standby load 8" is no longer showing up in the log.
Maybe it needed a short rest......who knows? So, of course a reboot fixed it 🙂
Thank you for your response!
I'll keep an eye on it.