Exchange Server - Store.exe problem

[placey]

Our exchange server has ceased working - \exhsrvr\bin\store.exe is taking 100% of the processor cycles - what's going on? has anyone else come accross this? I re-started the server but no change occures. Please help...

[patweb]

What conditions occurred leading up to this event? Are your harddrives filled up?

[placey]

Normal conditions seemed to prevail up to the time of 'failure'. hard drive has some 16Gig free. Machine is NT server 4.0 SP4 (yes I know about that but its been working fine for months now) with Exchange server 5.5 and proxy server 2.0 catering for 40 clients. No upgrades / changes of late. I run ESEUTIL on the databases off-line every two weeks or so. I am stuck!

[Equalizer]

Any errors running ESEUTIL?
Any errors in the event logs?
Which Exchange SP is installed?

Have you been through the KB article...

Troubleshooting High CPU Utilization By Store.exe (http://support.microsoft.com/support/kb/articles/q172/8/13.asp)

For instance...

ESEUTIL /D /ISPRIV

and

ISINTEG -pri -fix -test alltests

Run Gwclean and Profinst.

Apply the latest Service Pack/Hotfixes.

[This message has been edited by Equalizer (edited 08-18-2000).]

[patweb]

Did the problem occur following the ESEUTIL, or a server crash, or did it just shut down the STORE service and you are unable to get it back? Did you change or delete the Service account for the server?

[placey]

Thanks for the link Equalizer - this is a gold mine of information. To get the service running again I backup up then deleted \exchsrvr\mdbdata\* and (as I knew from experience) this forces Exchange to start its databases again. Store.exe settled and at least the users have email again. Email delivery for all clients is local machine so private email data loss should be minimal. Now I have to restore the public database - for us this is basically our public folders containing diaries / booking information. Although I have several backups of pub.edb they all seem to fail eseutil /g test - so I will go through and patch Exchange before trying again to retrieve our public database. I will post here the results.