SMGR Issues suddenly

[wpetilli]

began with an error at main web page to try logging in at another time. I restarted the cnd service, which fixed that. Navigating to the session manager dashboard generated an error page saying 'sorry, an unexpected failure has occured'. Navigating to routing/sip entities generated a message about

'One or more Session Manager elements may not be managed by this System Manager. The status could not be obtained. This may be due to a network connection problem or Session Manager elements older than release 6.3. For older Session Manager elements, no action is required as long as the Session Manager is configured to use this System Manager. If all Session Manager elements are not managed by this System Manager, administration changes should only be performed after understanding the implications of making administration changes on a System Manager that does not manage all Session Manager elements. Database changes on this System Manager are only replicated to the Session Manager elements that it manages.'

This has been working fine for many months. The ASM's are both licensed v7. I rebooted the server, which did not help. Ran a repair within replication and that also has done nothing. Has anyone seen this one before?? This is a smgr 7.0.1.3

 

[kyle555]

maybe tshark -i eth0 host "ip of a SM" from SMGR and see what happens why

If you've done something like say SMNetSetup and changed hostname but the replication password was expired, it's possible when SMGR hits up SM its looking for newhost and finding the IP in the hosts file, but the SM offers a cert issued to oldhost and SMGR sees a mismatch.

Something like that anyway, that management channel must be broken, and in working on another little mess yesterday came to the conclusion that enrolling in DRS is strictly dependent on the replication node's trusting of the SMGR CA. Either the trust relationship there isn't mutual cert exchange, OR the symmetric DS service doesn't actually care what cert the SM offers so long as it's signed by System Manager.

So, I got hostname newhost doing DRS fine, but can't be managed either and the cert for mgmt on the SM in inventory is issued to oldhost.

initTM -f if SMGR is your CA, or wireshark/pray/whatever.

 

[wpetilli]

Other than reboot nothing has been done with these in several months. Thought it could be something like that, but there were too many various issues not even related to ASM. Like just logging into SMGR. Some screens are fine, others weren't displaying properly. Weird

 

[wpetilli]

Other things.. going into replication an error stating 'geo not bound' along the top, but I can do a repair etc..

 

[wpetilli]

Looks like a hotfix per Avaya

 

[jimbojimbo]

'df -h' will indicate /opt is full. There is a file called "Avaya" which is probably over 2Gig. 7.0.1.3.1 fixes the issue. Quick fix is to stop JBoss, delete the file, 'touch' the file, change permissions back to match original file, restart JBoss. You will need to wait to restart JBoss for about 10 to 20 minutes until the actual disk space becomes available again ('df -h' should be around 65% on /opt).

 

[bankingguy]

looks like a bug bud.