Unity MWI stuck in 'pending'

[ken.montgomery]

So, we've been running a successful Unity deployment, integrated with Cisco Call manager for many years.  In the last few days, I've had a few issues with users whose voice mail lights are not synchronized.  Normally I just go in and turn off the MWI in Unity, and then turn it back on again, and everything is fine.

Today though, the user I'm having issues with won't fully re-enable.  I reset the MWI to be set for X (the extension of the phone) and save the user, when it returns it shows MWI request is pending ON. 

I have restarted unity, failed over to the secondary server, failed back, restarted the AVNofitiy service (spelling?) and tried deleting and re-adding the MWI a number of times... the only thing I see happening in event logs is the following:

SubSystem Message - Job 'VEY-UNITY-PRI-UnityDb-UnityDbPublication-VEY-UNITY-F-2' (0xBE8631C0A089294B9DD3A94A8959BBA3), step 2 - Login failed for user 'VOIP\VEY-UNITY-PRI$'. The process could not connect to Subscriber 'VEY-UNITY-F'.

So, I am wondering if I have a SQL problem or not? 

Anyone have any suggestions?  (I tried to open a TAC case, but of course, our SmartNet renewal has not made it through accounting yet.

[islam.kamal]

Dear

Yes you are completely true.When failover is configured, Cisco Unity uses Microsoft SQL Server replication to replicate data from the active server to the inactive server. If failover occurs, data replication ensures that current configuration and subscriber data is available on the secondary server and that, after failback, changes made on the secondary while it was active are replicated back to the primary. Replication is performed by SQL Server replication jobs, which are run by the SQLSERVERAGENT service.When SQL Server replication breaks, replication transactions are saved in audit log tables on the active server so the data can be replicated to the inactive server when replication is restored. If replication is broken for an extended period, the audit log tables can become large. This can cause performance degradation, which in turn can cause poor TUI response and can even cause failover to occur. Moreover, the Cisco Unity database on the inactive server does not have the latest configuration and subscriber data when it becomes the active server.

Please use the below link to solve your problem

http://www.cisco.com/en/US/products/sw/voicesw/ps2237/products_tech_note09186a008085ee56.shtml

Thank you

please rate if this will help

[ken.montgomery]

Thank you for your response.  While following through on this, even if failed over to the secondary server (which is also the exchange box) you cannot set the MWI.  Thus it isn't related to a replication issue, since we're all on the same server when I test it.

I also cannot find an error after going through the document you provide, as the accounts used are the proper ones.  I'll keep digging... if anyone else has other ideas, please let me know.

[David Hailey]

What version of Unity and Exchange are in use?

[ken.montgomery]

Those are important details I left out.

Cisco Call manager 6.1

Cisco Unity 5.1 (redundant Unity servers, one running exchange)

Exchange 2003

[ken.montgomery]

I gave up with no ideas over the weekend and no response from my consultant, so I deleted and recreated the mailbox and it is working.