I have a very interesting issue that I am working with Microsoft PSS on. We recently migrated from Exchange 2003 to 2007, as well as using Microsoft Data Protection Manager as our new back-up tool. The conscientious admin that I am, I like to do test restores to the RSG occasionally to make sure my back-ups are working right. Recently I started testing restores to the RSG with the new Data Protection Manager product. I was able to restore from D2D without any issue. I tried a restore from tape as well and was successful. However, later that night when the on-line maintanence ran I got a -1018 page mismatch error in my event log. I dismounted the RSG and ran eseutil /g on it and it failed with a -1206 database is corrupt message. Yikes! I removed the RSG and recreated it and restored from D2D and it was fine. Eseutil passed with no errors. Restored from tape and get the errors again running eseutil. Tried different tapes and have the same problem with errors running eseutil. Just to note, I have two storage groups and this issue is happening to both storage groups. I ran ntbackup to disk and restore to RSG with no problem...eseutil /g shows no errors. I then use my previous back-up product from EMC called Networker Exchange module. I do a back-up to tape, restore to the RSG, run eseutil and it passes with no errors. At this point it seems to me that Data Protection Manager has a flaw when restoring to the RSG from tape but not disk. I ran isinteg - test on my RSG after using the Networker from tape and DPM from disk and it found a bunch of warnings, but no errors. I even ran eseutil /k and eseutil /g on my production databases and they show no bad checksums, etc. Clean as a whistle. The Exchange PSS group says most likely DPM has an issue with tape restores to RSG, but they say I should run isinteg -fix on my production databases and try another restore to the RSG from tape, but I do not want to run -fix on databases that are probably fine. They also said I can move all my users to new storage groups but this would be a lot of work to me considering there is no proof the production databses are corrupt. I even have on-line check summing set-up when my on-line maintainence runs and I have no check sum errors so far. Can anyone shed any light on this? My databases/logs are on an HP EVA440 SAN, with fairly updated firmware. My tape drive is an HP LTO3 autoloader. Microsoft gave me a line saying my tape drive has not been tested with DPM and is not supported, although I can recover 65GB SQL databases and 100GB Virtual Machines with no problem.Thanks,Alex

Coruption in DB is mainly coz of corrupted Tables and Checksum isues.Run Eseutil /mh and make sure on the bottom of the result you dont have any "CHECKSUM ERROR FOUND"http://exchangeis.com/photos/articles/images/original/Eseutil-_2F00_MH-Database-Header.aspxRatish