Micros 3700 Database Validation Issues

[William M]

We are currently having some issues with our one locations restaurant's Micros POS system. For about a week and a half now, our end of night process has not been completing successfully. During this it backs up and attempts to validate the database. The database validation fails and the rest of the end of night process never completes. The purge historical totals step is not getting completed because of this leading to "XXXX number of checks have not been purged" messages on our terminals. As a temporary fix, I have scheduled the server to reboot each morning before opening so that the terminals function correctly.

This is what is contained in our C:\MICROS\Common\Etc\DM.log file where the error occurs:
VALIDATE TABLE "micros"."dsvc_dtl"
VALIDATE TABLE "micros"."dsvc_slu_def"
VALIDATE TABLE "micros"."dtl"
Table (dtl) has an incorrect page count (2014) in the catalog when (2015) was expected
SQL error (-301) -- Internal database error *** ERROR *** Assertion failed: 200610[Validate_MICROS] (11.0.1.2596)
Attempting to normalize a non-continued row (0x1054d:0x0) -- transaction rolled back
SQL error (-101) -- Not connected to a database
SQL error (-101) -- Not connected to a database
SQL error (-101) -- Not connected to a database
4 errors reported
8/29/2016 4:03:49 AM Database validation complete.
8/29/2016 4:03:50 AM Database validation failed, see \micros\common\etc\DM.log for details.
8/29/2016 4:03:50 AM DM Return Code = 3

If anyone has any knowledge on how to fix this incorrect page count error that causes validation to fail it would be greatly appreciated!

 

[pmegan]

You may want to run your backup as a windows scheduled task instead of through an autosequence. I do that to make sure the backup runs if the overnight aseq fails, but seems like it would help in this case as well.

As for fixing what you have going on, your best bet is to back up the database without validation, and then try to rebuild it in Database Manager. This will take a while, so try doing it late at night or very early in the morning.

 

[William M]

I will give that a try as soon as I get a chance to work after closing. Temporarily, I have edited the backup in the autosequence to just backup and not run the validation so the rest of our end of night actually completes and the historical totals are able to purge. Once I rebuild the database I will move the backup out to a scheduled task. That's sounds like a good plan to help with either issue.

After digging a little deeper into the cause of this issue, it seems that Windows Update ran the morning that this issue first started. For some reason, the person who set up this system has windows doing updates at 3am and our "End of Night" autosequence running at 3:30. The list of updates was a bit longer that day so it is possible that both of them were running at the same time and ended up corrupting something.

Thanks for the info!