AIR does not work after NBU appliance upgrade

[liuyang]

Hi, recently we helped a customer to upgrade two NBU 5230 appliance from 2.5.4 to 2.6.0.3. Before the upgrade, the AIR between the two appliance (one in prod site and one in DR site) is working fine. However, after upgrade, we can see AIR job run in the appliance in prod site. However, we cannot see any import job in the appliance in the DR site. 

My quesiton is: do I need to reset the AIR after upgrade? If no, how to troubleshoot in this case. Thanks in advance.

[chashock]

You shouldn't need to reset anything, no.  You should go through the steps in the AIR troubleshooting doc http://www.symantec.com/docs/TECH196435.

When you say you see the AIR job run on the prod side, I take it you mean the replication job, and that it completes with a status 0?

[liuyang]

Thanks for your reply. Yes, in the produciton site, it shows the backup and replication jobs completed successfully. However, in the DR site, it does not show anything. Normally we should see some import job running but after upgrade no import job is running in the DR site.

[chashock]

Do you get anything in the activity monitor?  Have you looked at the bpdm logs to see what errors might be showing up regarding the import?

[ejporter]

I had some similar problems in a 7.5 release.  Cycling nbstserv would restart the imports.   I never did chase it down as we were about to upgrade, and that resolved it.

You might try cycling nbstserv if you have not already done that or rebooted.

In either event I would suggest calling support, as you probably dont want to be checking you imports daily :)

[watsons · Accepted solution]

I take it that you did not see the import job at all, not that an import job appeared but failed (http://www.symantec.com/docs/TECH223243)

SLP settings has been integrated from file (LIFECYCLE_PARAMETER) into NBDB now. That means LIFECYCLE_PARAMETER is no longer used and deprecated.

Try to verify if those settings are still consistent in your new setting under "host properties -> SLP parameter", especially this parameter:

 MAX_MINUTES_TIL_FORCE_SMALL_DUPLICATION_JOB = <yourvalue>

A change of the value might prolong the import job to start. 

Some other possible causes:

1) Ensure timezone is not changed in target appliance

2) Check if you can see any event file in target appliance:  /disk/databases/spa/database/events/

There should be no file inside. If you find any replication event file, there is something wrong with the event manager not reading the file properly.

3) Check if there is any coredump of nbstserv or nbrmms, a core will definitely prevent import job from starting.

For #2 & #3, you will need to call up support for further investigation.