EV 9.0.1 server event id 3230

[Leonardo_Fern1]

I began to receive event 3230 about 24 hours ago. Access to my archives is ok, I followed the registry steps in article http://support.veritas.com/docs/285130 but the event keeps on being logged. Is the id referenced in the event an archive or my vault's id? Thank you.

Could not create a new MAPI session on Exchange server 'USATL01ME200'.
Please ensure the privileged mailbox has not been hidden, the server is running and that the Vault account has sufficient permissions on the server.

Internal References:
Unable to get exclusive access to the MAPI thread pool.
Mutex name: EnterpriseVaultProfileMutex 1B213F68A2F51CA44A31FEA1717A696281011z00AMERVAULT01
Reason: WAIT_TIMEOUT (10 mins)

[JesusWept3]

very strange, so what i would suggest is

1. Log on to the EV Server as the EV Admin account
2. Create a new Outlook profile to the system mailbox thats targeted by USATL01ME200
3. Attempt to connect to that mailbox, does it let you in without prompting for username and password?

a couple of things could be
1. Permissions aren't set correctly, make sure it has send as/receive as permissions, if it prompts you for username and password, the permissions have NOT been set correctly
2. Where is the system mailbox located? sometimes you might find someone has moved the system mailbox to a different exchange server, if thats the case, errors like this will occur, you need to move the system mailbox back.
3. It could be a Global Catalog issue, you could try using the Closest GC registry key detailed here:
http://support.microsoft.com/kb/319206
4. It may just need a simple Detect and Repair of outlook
5. If you have Outlook 2007, then you should install the performance hotfix released by Microsoft
http://support.microsoft.com/kb/2475891

[Leonardo_Fern1]

I logged into the EV server with the EVA account and there was no mapi profile for the ev system account. Only a vault agent profile. I created a profile for the ev system account and was able to open it with no problem. I haven't receivde a 3230 error i about 5 minutes so far.

[Leonardo_Fern1]

this didn't help :)

[JesusWept3]

hmm thats odd, it could have just been some MAPI inconsistency, keep monitoring it and hopefully it will just be chalked to being "one of those things"

[Leonardo_Fern1]

1 though 4 check out and I am running outlook 2003. 

[Leonardo_Fern1]

I read an article that mentions that ev servies shold be restarted weekly if not daily to prevent mapi errors. I've been using EV since 2006 and have never done this. All previous mapi eros were due to network problems, exchange servedr problems or even account lockouts. Is restarting ev servies daily or weekly still good practice? If so, can it be set via the EVA?

[JesusWept3]

OK then do the following for me

1. Stop the Task Controller service
2. Close Outlook
3. Go to Control Panel -> Mail -> Show Profiles
4. Delete each profile listed except for the one you created for the system mailbox
5. Do a Windows Search for "mapisvc.inf"
6. Delete the mapisvc.inf (some say copy the bigger inf over the smaller one, aimless excersise, just delete it, if you dont feel comfortable deleting it, just rename it)
7. Afterwards  go to Start -> Run -> type "FixMapi.exe" and press enter (you won't get any output, you won't even know its been run)
8. Now Open outlook, you may see that it says something like "Starting outlook for the first time", this is because its rebuilding the mapisvc.inf
9. Once outlook opens, start the the task controller
10. Go to the properties of the task in question and then go to the Synchronization tab and then synchronize a selected user and see if this works

The MAPI Troubleshooting sledgehammer works in most cases, if it still fails, then you either have an issue with the system mailbox being bad or on the wrong server (but you can connect in outlook and you checked out it IS hosted on the correct server, yeah?) or there may be something wrong with Outlook itself in which case you would need to do a detect and repair and possibly restart of the machine

If all else fails, can you stop all the tasks, start a dtrace of ArchiveTask and then start the task by itself so its the only one running and when the error occurs, stop dtracing, start all your other tasks and then post the dtrace here?

[Leonardo_Fern1 · Accepted solution]

Reboot fix it. I did the old fashion thing and simply rebooted the box last night and the mapi errors are gone. I actually hate doing that because I still don't know what really caused the mapi failures. Thanks for all your help though.

[JesusWept3]

good job :)
don't forget to mark it as solved!