Enterprise vaule 6.0 index is failing to rebuild help needed.

[andrew_charles]

Hi,
Our Enterprise vault which is 6.0 service pack 2 i think. Has a index that has failed. When i try to rebuild it it just stays on the rebuilding faze. it is throwing up errors that are event ID 3277, 2270 and 41021. There is another thred on here that pointed me to the index failing. But am wondering how i can find out why the index keeps failing to re-build? and how i can get the index to stay in one piece?

Many thanks in advance

Charlie

[Liam_Finn1]

 First why are you rebuilding the index

If you have backups before the failure stop the index service, restore the last known good index from backup then start the services and update the indexes

[Maverik]

Failing what Liam said can you post the whole events not just the codes.

[Liam_Finn1]

 Also do a DTrace  in verbose mode of the Indexbroker and Storagebroker so we can see what is happening in EV when the process fails

[andrew_charles]

Ok well started getting errors on the vault server. Upon doing some research on here i discovered that one of the causes for the errors is that the Index has failed. I checked the index using the IndexVolumeRelay.exe it had failed. So i started to rebuild it. It seems to take ages then says that it has been submited ok. It then starts to rebuild and does not seem to rebuild/update. We had a guy from Blue source on the server the other day looking at it. When i seen him rebuild the archive it took like 3 minutes to re-build. Am just wondering why when am trying to doit now its not rebuilding. the first error that i get is:

Type: error   Event ID 3277

The MAPI error MAPI_E_FAILONEPROVIDER (0x8004011D) has occurred whilst synchronizing the properties of mailbox /o=StyleGroup/ou=First Administrative Group/cn=Recipients/cn=stanifs.

The mailbox will not be synchronized until the next scheduled run.

.\AgentExchSynch.cpp (CAgentExchSynch::SynchronizeMailboxEntry) [lines {585,635}] built Feb 26 23:02:25 2006

then i get the following warning

Type: Warning    Event ID: 2270

A queued operation exceeded the retry count and has been discarded



m_pIExchangeSynch->SynchronizeMailboxEntryEx(LegacyMailboxDn = "/o=StyleGroup/ou=First Administrative Group/cn=Recipients/cn=stanifs",
 adMbxDn = "CN=Staniforth\, Steve,OU=Service Engineers,OU=!_Mobile Users,OU=Bradford,DC=stylegroup,DC=local",
 adMsgStoreDn = "CN=Mailbox Store (BRA-MAIL2),CN=First Storage Group,CN=InformationStore,CN=BRA-MAIL2,CN=Servers,CN=First Administrative Group,CN=Administrative Groups,CN=StyleGroup,CN=Microsoft Exchange,CN=Services,CN=Configuration,DC=stylegroup,DC=local",
 SynchModes = "3")
HRESULT: 0xC0040CCD

Cant run Dtrace with the index rebuilding for some reason though

[andrew_charles]

Also have just checked the backups and the index is not backed up. Looks like who ever set this up thought there could be better use for the space. Not happy soon as this is sorted will bhe setting it to back up. So any help with this is going to be great. Oh and there is a monitoring error also.

There are 100286 savesets that are waiting to be indexed and/or waiting to be backed up or replicated in Vault Store 'Journal Store'.

Check that your Indexing and Storage services are running.

Review your procedures and make any changes needed to ensure that backups happen in a timely fashion.

If you are using an EMC Centera, check the size of the Centera replication queue.

For more information, see Help and Support Center at

This is the error that led me to rebuild the index.

[Liam_Finn1]

The second warning is that there is a backlog of items waiting to be written to an index and the index is not available. Maybe it is the index you are having issue with 

[Liam_Finn1]

 This may sound like too simple a fix. Restart the EV services on that server

[andrew_charles]

Right that seems to have worked. Restarted the services. Left it for 10 mins to settle down and is now archiving away merry as can be. Thanks for the help scanner001. Quick one though how do i force the backlog to be written to the index? Do i have to wait for the index to repair itself?

[Liam_Finn1 · Accepted solution]

 Yes if it is writing to that index then yes you need to wait for the repair to finish. If after the repair is finished and the backlog does not drop then you can restart the storage service to kick start the write. Don't restart the storage service until the rebuild is finished

If this solved the issue for you please make the appropriate post as the solution

Liam