Exchange Server Forums

Raid Arrary Drive Replaced - Errors

Users viewing this topic: none

Logged in as: Guest

Printable Version

All Forums >> [Microsoft Exchange 2003] >> Information Stores >> Raid Arrary Drive Replaced - Errors

Page: [1]

Message

<< Older Topic Newer Topic >>

Raid Arrary Drive Replaced - Errors - 8.Feb.2012 1:50:48 AM

slarabee

Posts: 4
Joined: 8.Feb.2012
Status: offline

Hello All,

I am running Exchange 2003 and had a drive stop responding in a SATA SW Raid 5 array on Server 2003 on yesterday.

We pulled the drive and reseated and the drive came back online and rebuilt the data and page and everything is fault tolerant again and healthy.

However when I ran my online backup i received error:

' from a call to 'HrESEBackupRead()' additional data '-'The 'Microsoft Information Store' returned 'Error returned from an ESE function call (d).
' from a call to 'HrESEBackupRead()' additional data '-'

So after some checking here I see there may be some corruption in priv1.edb that needs to be addressed.

Now the good news is that the Information Store is mounted and seems to be working fine (for now, knock on wood) but I intend to address the issue Wednesday night and get everything cleaned up so I could use a little nudging in the right direction as to what to do first.

I have a good back up (I hope knock on wood) that is just a couple of days old and the journal running copies of all emails sent and received.

As I say the server appears to be running normally so:

Should I start with running ISINTEG and see what errors I get?

Then run ESEUTIL soft recovery and offline defrag?

As I say any help on where to start would be great.

Thanks,

Sean

< Message edited by slarabee -- 14.Feb.2012 12:48:31 PM >

Post #: 1

Featured Links*

RE: Backup Error - Advice Needed for Starting Point - 9.Feb.2012 10:55:45 AM

slarabee

Posts: 4
Joined: 8.Feb.2012
Status: offline

Just an update in case anyone else ends up in this situation.

I ran isinteg -s servername -test alltests and it found no errors and 3 warnings.

I ran isinteg -s servername -fix -test alltests and it did not find any errors and fixed the three warnings.

I rebooted and reran my online backup and it got to the point where it had been failing (about 12gb in) and hung...

While I was looking at all the ramifications of running the eseutil it came out of hanging and the backup was running well past where it had previously failed.

So I sat back and let the backup run. Surprisingly it finished and checking the report appeared to have no errors.

I checked the Application logs and found:

Information Store (3600) First Storage Group: A read operation on the file "E:\Program Files\Exchsrvr\mdbdata\pub1.edb" at offset 600768512 (0x0000000023cf0000) for 65536 (0x00010000) bytes failed 1 times over an interval of 8.672 seconds before finally succeeding. More specific information on these failures was reported previously. Transient failures such as these can be a precursor to a catastrophic failure in the storage subsystem containing this file.

So being that I did not want to work on it all night I decided that before running any further utilities on the data I would run another online backup tonight and see how it goes. If I do not get any ESE errors I will be happy. :-)

< Message edited by slarabee -- 9.Feb.2012 11:05:19 AM >

(in reply to slarabee)

Post #: 2

RE: Backup Error - Advice Needed for Starting Point - 11.Feb.2012 12:43:24 PM

slarabee

Posts: 4
Joined: 8.Feb.2012
Status: offline

It appears that I am talking to myself here lol but I will post replies just the same as maybe they will help someone in a similar situation.

Turns out the corruption is in pub1.edb (as is obvious in the error in the reply above) so I need to repair that data. Feel kinda silly that I missed it but both the stores mount fine and have had no problems outside of backup failure and ese errors in the application logs.

I have a backup to work from if I need to restore the data but... For now I am going to try to run eseutil /r to see if I can fix it that way. The only concern I have is that my priv1.edb is clean and I do not want to cause any problems there and I cannot see a way to run eseutil /r ONLY on pub1.edb but I am still reading...

Does running eseutil /r seem like a good idea in this scenario? Well Sean I am not sure but you can give it a try. Why thank you I think I will. Let me know how it goes? I will do that, thank you so much for your interest. :-)

< Message edited by slarabee -- 11.Feb.2012 12:46:12 PM >

(in reply to slarabee)

Post #: 3

RE: Backup Error - Advice Needed for Starting Point - 14.Feb.2012 12:44:45 PM

slarabee

Posts: 4
Joined: 8.Feb.2012
Status: offline

Conclusion:

Well it seems that so far no one was interested in commenting or offering any advice on the issue I was trying to resolve. That's okay busy lives of busy people lol :-)

Still I thought I would post how the issue was resolved in the end and what I think I did wrong.

First the solution:

I tried every repair I could think of but in the end the repairs were failing and defrag /p and /r would not complete successfully giving me jet read write errors.

I was about to overwrite the data with my last on line backup and restore etc... when I thought I would try the /i switch and ignore the streaming file mismatch check during the repair.

Viola!!

So what I ran to fix the issue:

eseutil /p "e:\exchsrvr\mdbdata\pub1.edb /i
eseutil /p "e:\exchsrvr\mdbdata\priv1.edb /i

eseutil /d "e:\exchsrvr\mdbdata\pub1.edb
eseutil /d "e:\exchsrvr\mdbdata\priv1.edb

isinteg -s servername -fix -test alltests (on both the public and the private stores)

From http://exchangeis.com/blogs/exchangeis/archive/2007/03/29/eseutil-p-a-quick-and-easy-tutorial.aspx

ESEUTIL /p <database name> [options]
/s<file>     Streaming file location (Optional default is not to use)
/t<db>      Temporary Database Location and name default is: TEMPREPAIR*.EDB
/f<name>    Prefix for database report files, the default is: <database>.integ.raw)
/i           bypasses the mismatch error check on database and streaming file
/g           To run the integrity check before repairing database
/createstm   Creates and empty stm streaming file if this is missing
/8           To set the 8k database page size (default: auto-detect)
/o           suppress logo

Lastly what I think I did wrong was kind of a stupid mistake.
When I replaced the drive I did not set the stores not to mount on start up. I think I should have disabled the MS Exchange Information Store Services. Dismounted the stores and unchecked mount at start up and I could have saved myself the headache.
I hope this helps anyone else in this situation.

< Message edited by slarabee -- 14.Feb.2012 12:46:27 PM >

(in reply to slarabee)

Post #: 4

RE: Backup Error - Advice Needed for Starting Point - 16.Feb.2012 10:53:19 AM

nelson

Posts: 39
Joined: 5.Dec.2010
Status: offline

Thanks for sharing your solution. I have read many articles and most of them recommend not to use eseutil repair utility first because after running these tools, user may have to face the data loss situations and sometimes user may loss his important data. It is last sort. However, there are some third party Exchange server recovery applications which can help the users to avoid such data loss issues.

(in reply to slarabee)

Post #: 5