Don’t you just love it when a client’s exchange server decides to lockup and then fail to mount when it comes back up. Even better when it happens 1 hour before knockoff time…

Well this one had me stumped for a while. It is quite a large database (for SBS 2003)  edb+stm = around 60Gb

Here are the steps that I looked at to get it working again.

First things first – check in the event logs…

Heaps of

[message_box title=”MSExchangeSA Event ID: 9175″ color=”red”]The MAPI call ‘OpenMsgStore’ failed with the following error: The Microsoft Exchange Server computer is not available. Either there are network problems or the Microsoft Exchange Server computer is down for maintenance. The MAPI provider failed. Microsoft Exchange Server Information Store[/message_box]

and for every user that uses activesync to sync iPads/iPhones etc…

[message_box title=”Server ActiveSync Event ID: 3005″ icon=”alert.png”]Unexpected Exchange mailbox Server error: Server: [server.fqdn] User: [user@fqdn] HTTP status code: [503]. Verify that the Exchange mailbox Server is working correctly. [/message_box]

I tried to Mount the Private store and this error appears…

[message_box title=”MSExchangeIS Event ID: 9518″ icon=”alert.png”]Error Out of memory starting Storage Group /DC=local/DC=dcname/CN=Configuration/CN=Services/CN=Microsoft Exchange/CN=domain/CN=Administrative Groups/CN=first administrative group/CN=Servers/CN=dcname/CN=InformationStore/CN=First Storage Group on the Microsoft Exchange Information Store. Storage Group – Initialization of Jet failed. [/message_box]

I check the consistency of theexchange databases using…

“ESEUTIL /mh priv1.edb”

Dirty Shutdown – OK. There were some un-played log files so I performed a soft recovery which was successful.

An example of this would be “eseutil /r e00 /l [log_path] /d [database_path]”
In this example the logs start with e00. If your logs and database are in different locations you can use the /l  and /d parameters.

Clean Shutdown being reported now – great! Still getting 9518 on attempting to mount store. By this stage I knew I had ShadowProtect Backups of the entire server so I decided to recover the exchange from a short while before it locked up.

I stopped the Information Store service. Moved the existing store to another folder and recovered the entire exchange (which sits on a second drive – not the system drive) folder with .stm, .edb, .chk  and .log files. If your checkpoint (.chk) files are still located on the C:\ drive you may need to recover those too.

Tried to mount store – failed. Removed checkpoint files, replayed log files. Tried to mount store – same Event ID 9518. Back to the drawing board.

Time to try a defrag on the databases.

eseutil /ms “priv1.edb”

eseutil /ms “pub1.edb”

Both successful. Attempt mount again – failed event ID: 9518. OK – something else going on here. I don’t really want to perform a hard recovery if I don’t have to. The databases actually seem to have pretty good integrity.

This server was running exchange 2003 SP2. So I downloaded the Service Pack again and ran it over the existing install. If your server is only running SP1 then probably run that.  It asked for a reboot.

Upon reboot – guess what.  Log files replayed and Stores are mounted!

The Moral of the Story…

Never give up.

Even if you can’t find a solution on Google (which I couldn’t in this case) try to think outside the box.

Remember – There is always Microsoft PSS if you are really stuck.