edd_t Posted July 31, 2007 Share Posted July 31, 2007 One of our public information stores is corrupt, so we cant mount the database. I've got a ton of angy uses with out e-mail! ahhhhhh. When I mount the store I get the following error: The database files in this store are corruped. ID no: c104173b Exchange System Manager The knowledge base doesnt seem to have info on that ID number! In the even viewer I get this: Information Store (4708) First Storage Group: The database page read from the file "M:\EXCHSRVR\mdbdata\priv1.edb" at offset 26745671680 (0x000000063a2a9000) (database page 6529704 (0x63A2A8)) for 4096 (0x00001000) bytes failed verification due to a page checksum mismatch. The expected checksum was 3949998620039364478 (0x36d136d122babf7e) and the actual checksum was 8189920891379784918 (0x71a871a89d3e1cd6). The read operation will fail with error -1018 (0xfffffc06). If this condition persists then please restore the database from a previous backup. This problem is likely due to faulty hardware. Please contact your hardware vendor for further assistance diagnosing the problem. Have run a checksum and the results are so far: C:\Program Files\Exchsrvr\bin>eseutil /k "m:\exchsrvr\mdbdata\priv1.edb" Microsoft® Exchange Server Database Utilities Version 6.5 Copyright © Microsoft Corporation. All Rights Reserved. Initiating CHECKSUM mode... Database: m:\exchsrvr\mdbdata\priv1.edb Streaming File: m:\exchsrvr\mdbdata\priv1.STM Temp. Database: TEMPCHKSUM2604.EDB File: m:\exchsrvr\mdbdata\priv1.edb Checksum Status (% complete) 0 10 20 30 40 50 60 70 80 90 100 |----|----|----|----|----|----|----|----|----|----| .........................ERROR: page 6529704 checksum failed ( 0x36d13 6d122babf7e / 0x71a871a89d3e1cd6 ) .......................... 12722034 pages seen 1 bad checksums 0 correctable checksums 20162 uninitialized pages 0 wrong page numbers 0x72af9f3f highest dbtime (pgno 0x1a44fd) 795128 reads performed 49695 MB read 495 seconds taken 100 MB/second 506757043 milliseconds used 637 milliseconds per read 153079 milliseconds for the slowest read 0 milliseconds for the fastest read File: m:\exchsrvr\mdbdata\priv1.STM Checksum Status (% complete) 0 10 20 30 40 50 60 70 80 90 100 |----|----|----|----|----|----|----|----|----|----| ................................................... 4842496 pages seen 0 bad checksums 1656585 uninitialized pages Operation terminated with error -1206 (JET_errDatabaseCorrupted, Non database fi le or corrupted db) after 674.766 seconds. C:\Program Files\Exchsrvr\bin> Anyone got any ideas? I really don't want to do a restore... Quote Link to comment Share on other sites More sharing options...
Ark Posted July 31, 2007 Share Posted July 31, 2007 Break out the credit card and call MS. Quote Link to comment Share on other sites More sharing options...
robin Posted July 31, 2007 Share Posted July 31, 2007 Can you restore the file from a backup. People may lose a few emails but at least it might mount the database OK. Any ideas on how the file got corrupted? It may just do it again if the disk or memory is flakey. Quote Link to comment Share on other sites More sharing options...
ivan Posted July 31, 2007 Share Posted July 31, 2007 Now that fully justifies why I never used MS Exchange. Quote Link to comment Share on other sites More sharing options...
edd_t Posted July 31, 2007 Author Share Posted July 31, 2007 Can you restore the file from a backup. People may lose a few emails but at least it might mount the database OK. Any ideas on how the file got corrupted? It may just do it again if the disk or memory is flakey. i could restore from backup, i've save the transaction files so not many e-mails would be lost. but first of all im running: eseutil /p priv1.edb so its running repair mode but still scanning the database. ahhhhhhh its gonna take ages! not sure how it got corrupted. hardware seems to be fine. but i'll do a full investigation into that tommorow. Quote Link to comment Share on other sites More sharing options...
Jake Posted July 31, 2007 Share Posted July 31, 2007 Have you tried turning it off and on again? Quote Link to comment Share on other sites More sharing options...
suprafan72 Posted July 31, 2007 Share Posted July 31, 2007 Are u running exchange on a single instance or a clustered node? Quote Link to comment Share on other sites More sharing options...
edd_t Posted July 31, 2007 Author Share Posted July 31, 2007 Are u running exchange on a single instance or a clustered node? clusterd. i tried moving the services to the other cluster but no joy. Quote Link to comment Share on other sites More sharing options...
edd_t Posted July 31, 2007 Author Share Posted July 31, 2007 Have you tried turning it off and on again? haha thats what buggerd it up in the first place Quote Link to comment Share on other sites More sharing options...
T2 MSW Posted July 31, 2007 Share Posted July 31, 2007 Ive had this happen before on a Cluster. In our instance it was due to the Q drive being unpresented from the SAN by the storage manager. We eneded up restoring from the last backup. I would recommendation restoration from the last backup especially if you have the transaction logs. The hard recovery can sometimes be a little too harsh and you will prob end up with loads of users that can see the summary of mail items but when they go to get the full item they get an error due to the data being missing. Quote Link to comment Share on other sites More sharing options...
edd_t Posted July 31, 2007 Author Share Posted July 31, 2007 oh no dont say that im just playing the waiting game at the moment for this eseutil restore. Quote Link to comment Share on other sites More sharing options...
Schtuv Posted July 31, 2007 Share Posted July 31, 2007 Do you replicate the data on the PFs anywhere else? I lost a PF store recently - just deleted the old one, started the store (to create a new blank), and all the data arrived back overnight. Quote Link to comment Share on other sites More sharing options...
Schtuv Posted July 31, 2007 Share Posted July 31, 2007 Looking at the event log entry my moneys on a failed disk write (bad disk or high queuing causing a discard) causing the corruption ... BTW Microsoft have some new Exchange Server performance analysis tools out, so give them a run and see what reports back. EDIT: I see you've run eseutil /k ... have you tried the /r option, yet (with a log prefix, and /i)? Get a flat file backup of the database and logs first. Quote Link to comment Share on other sites More sharing options...
GeordieSteve Posted July 31, 2007 Share Posted July 31, 2007 What you using for backups mate? Quote Link to comment Share on other sites More sharing options...
Schtuv Posted July 31, 2007 Share Posted July 31, 2007 One of our public information stores is corrupt, so we cant mount the database. I've got a ton of angy uses with out e-mail! ahhhhhh. Hang about - are you saying that your users store and PF store are in the same storage group, and because the PF store is corupt, the users store won't start? Quote Link to comment Share on other sites More sharing options...
edd_t Posted July 31, 2007 Author Share Posted July 31, 2007 Hang about - are you saying that your users store and PF store are in the same storage group, and because the PF store is corupt, the users store won't start? oops no. meant to put private store there... all is working again now. cant remeber the exactly what it said but something to do with the jet database thingy! will check out the new analysing tools tommorow as no doubt someone will wanna know what happend. im guessing its gonna be a disk fail due to high queing like u said. we actually backup the exhange with 2 systems at the moment. the main backup is Backup Exec 11d. But we have started using a HP DS Client online backup which is still in testing mode. Quote Link to comment Share on other sites More sharing options...
Schtuv Posted July 31, 2007 Share Posted July 31, 2007 all is working again now. cant remeber the exactly what it said but something to do with the jet database thingy! will check out the new analysing tools tommorow as no doubt someone will wanna know what happend. im guessing its gonna be a disk fail due to high queing like u said. Exactly the same error here, on 2 privs and one pub.edb, on valentines day (typical). I worked a 45 hour shift to get the system back Micro$oft recommended migrating everything to new clean databases (just using 'move' in ESM) 'just in case', as they don't recommend using stores which have been in that state in production. That did isolate the problem mailbox(es), though. Turned out that somehow we had most of the logs and databases running off one drive (NOT how I config'ed it originally!) which killed the disk throughput. And an av trying to scan the edbs. *sigh* EDIT: Should we try to somehow relate this to Supras, given the nature of this forum? "You know, Paul, running an Exchange cluster is much like driving a supra ..." Quote Link to comment Share on other sites More sharing options...
MarkR Posted July 31, 2007 Share Posted July 31, 2007 Try keep the size down.. helps when it comes time for eseutil Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.