"Moving the directory around" would not fix the -1018 error because moving
the database to a different drive will not fix corruption. The -1018 errors
are usually caused by faulty hardware, but poor administrative practices can
help these occur, for instance, not properly shutting down servers before
turning off the power.
Ed Crowley MCSE+Internet MVP
Tech Consultant
Compaq Computer Corporation
All your base are belong to us.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]On Behalf Of Semiglia,
Michael
Sent: Wednesday, August 29, 2001 9:42 AM
To: Exchange Discussions
Subject: Restore of directory.
Hi all,
Exchange 5.5 on NT 4. Veritas Backup Exec and TSM/ADSM backup solution.
We had some issues recently where we had a corrupted directory on a server
and I had a brain cramp in implementing the restore. Looking back, it was
definitely an oversight on my part, however even after the 20/20 hindsight,
I see no easy way around the problem.
We were doing service pack updates on all of our servers from NT SP4 to NT
SP6a and then from Exchange SP2 to SP4. Prior to packing the servers I had
an online and offline back available. This was server number 18 out of 19.
Work had been commencing over a 3 month period and all our servers are on
Compaq proliants with the same h/w and s/w configuration. Installed both NT
SP6a and Exchange SP4 on this particular server with no unusual entries in
the event logs. However on the next backup of the directory using backup
exec, started getting the following -1018 errors:
Event Type: Error
Event Source: ESE97
Event Category: Database Page Cache
Event ID: 118
Date: 8/12/2001
Time: 11:50:27 AM
User: N/A
Computer: CAEEXP01
Description:
MSExchangeDS (471) Direct read found corrupted page (25075) with error
-1018. Please restore the databases from a previous backup.
The indication in technet was that there is a hardware issue. At that point,
using perfopt, moved the directory around to different physical/logical hard
drives. Also ran the eseutil with the verify switch. Moving the directory
around still caused the above error and the eseutil program returned no
errors.
At this point I tried using the TSM/ADSM backup solution that we were
rolling out and this caused the same -1018 error. This pretty much told me
that there was no issue with the backup solution and sort of exonerated the
h/w.
At this point I decided to do a restore of the directory from the last good
backup I had, which was the night prior to the service pack work. I placed
the tape in and kicked of the restore. Now for all of you reading along,
this is where you go "You can't do that". The braincramp is that the server
is at NT SP6a and Exchange SP4 now and the backup set is at NT SP4 and
Exchange SP2. The disaster recovery whitepaper specifically states that the
service pack levels need to be the same as the backed up copy. I also
realized that after I kicked of the restore. The restore finished
successfully and all I need to do was start the directory. At this point, I
said just let me see what happens. I had an online and offline backup of the
corrupted directory and could return to the point I was prior to doing the
restore. After starting the directory it started replaying log files and the
came up with the following error:
Event Type: Error
Event Source: ESE97
Event Category: Logging/Recovery
Event ID: 134
Date: 8/25/2001
Time: 2:22:18 PM
User: N/A
Computer: CAEEXP01
Description:
(501) The database s:\exchsrvr\DSADATA\dir.edb created at 10/22/1997
20:51:54 was not recovered.
After that, I got a bunch of table rebuilds indicating that something was
aware that the restored directory was NT SP4 and the OS was now NT SP6a.
After that everything came up smoothly. No errors anywhere in sight. I then
performed a backup and it worked w/o and errors on both backup exec and the
TSM/ADSM backups.
I really don't understand the above error. Anybody have any insights on that
one?
I guess my question is, is the directory all right or should I run as fast
as I can to the server an turned it off before it poisons my entire
organization.
Thank You!
Michael Semiglia
Exchange E-mail Engineering
*207 575-1094
*[EMAIL PROTECTED]
_________________________________________________________________
List posting FAQ: http://www.swinc.com/resource/exch_faq.htm
Archives: http://www.swynk.com/sitesearch/search.asp
To unsubscribe: mailto:[EMAIL PROTECTED]
Exchange List admin: [EMAIL PROTECTED]
_________________________________________________________________
List posting FAQ: http://www.swinc.com/resource/exch_faq.htm
Archives: http://www.swynk.com/sitesearch/search.asp
To unsubscribe: mailto:[EMAIL PROTECTED]
Exchange List admin: [EMAIL PROTECTED]
