We have an Exhange server in our organisation running Exchange 5.5 SP4 on an NT4 SP6a server.
We ran out of space on the drive where the databases are (a planned upgrade didn't happen in time). This caused the MTA to stop.
We added in another disk, expanded the array and extended the partition.
After this we ran
ESEUTIL -P /ISPRIV
ESEUTIL -P /ISPUB
Deleted the edb.chk files
Started the DS
Ran ISINTEG -patch (no inconsistencies found)
Exchange was then started and, following a reboot, messages began to be processed. There are no messages sitting in the MTA queue. However, messages in users outboxes from when the system was down have remained in their outboxes. Doing a Send and Receive doesn't clear them. The Outboxes cannot be accessed to delete the messages.
A few users have problems whereby they cannot access certain messages or folders. One user cannot access their sent items folder, another has an inbox which reports 35 unread messages, however there aren't any messages in the folder. They are able to see new messages coming in, but haven't received the ones that were sent whilst the system was down.
In the MTADATA folder there are 5 large db*.dat files, one of which is 1.2Gb in size. In the application Event log the following errors are being reported several times every second:
EVENT: 2110
SOURCE: MSEXCHANGEMTA
Category: Internal Processing
A fatal MTA database server error was encountered. A bad list member length is on object 06000047. File offset: 1330. Attribute ID: 79. Referenced object 00000000 (0 => N/A). Referenced object error: 0. [DB Server DISP:ROUTER 10 42] (16)
There is also the associated EVENT 2171 warning message.
So, we re-ran the ESEUTIL ISINTEG process described above (twice for good measure) and then started all services except the MTA and ran MTACHEK /rd /rp/ /v/ /f mtacheck.log
No errors were returned, all reported as ok.
The MTA db files are still large and growing. The MTA errors are still in the application log and the users still have the same problems with their mailboxes.
My next step is to delete all the db*.dat files from the MTADATA and replace them with the default ones from the Exchange CD. I realise that this will result in some data being lost.
Does anyone have any other ideas? Do you think this will resolve the issue with the users mailboxes or do you suspect (as I do) that we will have to restore these from backup?
All advice gratefully received.
We ran out of space on the drive where the databases are (a planned upgrade didn't happen in time). This caused the MTA to stop.
We added in another disk, expanded the array and extended the partition.
After this we ran
ESEUTIL -P /ISPRIV
ESEUTIL -P /ISPUB
Deleted the edb.chk files
Started the DS
Ran ISINTEG -patch (no inconsistencies found)
Exchange was then started and, following a reboot, messages began to be processed. There are no messages sitting in the MTA queue. However, messages in users outboxes from when the system was down have remained in their outboxes. Doing a Send and Receive doesn't clear them. The Outboxes cannot be accessed to delete the messages.
A few users have problems whereby they cannot access certain messages or folders. One user cannot access their sent items folder, another has an inbox which reports 35 unread messages, however there aren't any messages in the folder. They are able to see new messages coming in, but haven't received the ones that were sent whilst the system was down.
In the MTADATA folder there are 5 large db*.dat files, one of which is 1.2Gb in size. In the application Event log the following errors are being reported several times every second:
EVENT: 2110
SOURCE: MSEXCHANGEMTA
Category: Internal Processing
A fatal MTA database server error was encountered. A bad list member length is on object 06000047. File offset: 1330. Attribute ID: 79. Referenced object 00000000 (0 => N/A). Referenced object error: 0. [DB Server DISP:ROUTER 10 42] (16)
There is also the associated EVENT 2171 warning message.
So, we re-ran the ESEUTIL ISINTEG process described above (twice for good measure) and then started all services except the MTA and ran MTACHEK /rd /rp/ /v/ /f mtacheck.log
No errors were returned, all reported as ok.
The MTA db files are still large and growing. The MTA errors are still in the application log and the users still have the same problems with their mailboxes.
My next step is to delete all the db*.dat files from the MTADATA and replace them with the default ones from the Exchange CD. I realise that this will result in some data being lost.
Does anyone have any other ideas? Do you think this will resolve the issue with the users mailboxes or do you suspect (as I do) that we will have to restore these from backup?
All advice gratefully received.
