Exchange 5.5 errors and failure

[mleggatt]

This evening our Exchange 5.5 SP3 server failed. The actual NT server was still runing but none of the Exchange services. I checked the event log and found the following error messages, some of them more than once:

Event Type: Error
Event Source: ESE97
Event Category: Logging/Recovery
Event ID: 126
Date: 22/07/2003
Time: 21:49:30
User: N/A
Computer: *******
Description:
MSExchangeIS (255) Unable to write to section 0 while flushing the log. Error -1022.

Event Type: Error
Event Source: NTBackup
Event Category: None
Event ID: 8010
Date: 22/07/2003
Time: 21:49:43
User: N/A
Computer: *******
Description:
Microsoft Exchange services returned 'The remote procedure call failed.
' from a call to 'BackupRead()' additional data ' - '

Event Type: Warning
Event Source: MSExchangeMTA
Event Category: Internal Processing
Event ID: 2181
Date: 22/07/2003
Time: 21:49:54
User: N/A
Computer: *******
Description:
An MTA database server error was encountered while securing an object. Called from MTA. Procedure 101. Database error code: 1117. Object at fault: 060000B6. [DB Server XFER-IN 13 38] (14)

Event Type: Error
Event Source: MSExchangeMTA
Event Category: Internal Processing
Event ID: 2026
Date: 22/07/2003
Time: 21:49:56
User: N/A
Computer: *******
Description:
An internal MTA error occurred. Contact Microsoft Technical Support. Unable to write to the message tracking log, error '. [BASE XFER-IN 8 40] (16)

Event Type: Error
Event Source: MSExchangeSA
Event Category: Monitoring
Event ID: 1005
Date: 22/07/2003
Time: 21:54:49
User: N/A
Computer: ********
Description:
Unexpected error <<0xc0040000 - Network problems are preventing connection to the Microsoft Exchange Server computer. Microsoft Exchange Server Information Store ID no: 80040115-0514-000006bf>> occurred.

This server has been know to have problems with its SCSI card previosuly and just rebooting the server resolved it. I did the same this time and the server and Exchange services restarted without any errors in the event log. From what I have read the initial error is because of a corrupt EDB.LOG file which caused the other services to fail and hence the other errors. I think the reboot means that the EDB.LOG file was checked and corrected.

My questions are:

1. Is this indicative of possible future faults? i.e. corruption in databases

2. Is there a safe method for checking Exchange databases for consistency i.e. the ISINTEG utility?

I intend to read more myself but any hints/assistance would be appreciated.

Thanks in advance

 

[PaulEssick]

I run ISINTEG and ESEUTIL monthly on our servers. I recommend doing a full backup just before running them. There are several switches that can be set, you need to decide which ones you want to use.

The problem with the log file could be caused by data corruption, a bad spot on the hard drive, or any number of other things. When I have the downtime I like to do hardware checks to make sure they pass all of the tests.

Good luck,
Paul

 

[fuego007]

In my opinion, running the isinteg and/or eseutil on a regular basis is a waste of time. In addition, Microsoft strongly recommends against using the eseutil as a tool for performing routine maintenance. Eseutil should only be used to resolve problems that are preventing the information store from starting (or that have caused it to stop). Isinteg can be used to help resolve problems with the private and public information store databases, such as corruption, but should only be used if you suspect something is wrong.

Your databases MUST be consistent for them to start.

If you are experiencing hardware issues, such as reliability, replace the server ASAP. It's MUCH cheaper than losing a lot of e-mail and being without e-mail service for several days.

I have a FAQ for replacing an existing server with new hardware, should you decide to do this.

Bob

 

[mleggatt]

Thanks for the post guys, all useful info. I agree with fuego this server needs replacing. It failed once last night and has failed at least four times previously that I know of each time reseating the SCSI cable has resolved the issue. There is a definite fault with the built in SCSI controllers on the motherboard. The server is about 5 years old and showing it's age.

We are in the process of planning our upgrade to Windows 2000 servers and Active directory but as yet have not done this. If I had my way I would upgrade to Exchange 2000 ASAP and migrate all the mailboxes off this server onto the new one. Trouble is we need the Active directory domain before we do this. I also have the problem that my peers seem to think well it works and you can keep it going so why worry!!!! I think I need to highlight the possible down time if this server fails completely!!

Thanks again guys