Administration Guide : Xinet Statistics and Database Operations — the Database, Admin subtab : Scheduling Table Checks

Scheduling Table Checks
Logging information for Table Checks is kept in the file, /var/adm/appletalk/check.log on Unix systems and C:\PROGRA~1\Xinet\FULLPR~1\check.log on Windows systems.
Important: Schedule your Solr Database purge process to remove unwanted index entries of deleted files before scheduling a backup or periodic sync and table checks to avoid CPU performance issues. For more information, see Setting your own Schedule to Purge Deleted Files From your Solr Database.
 
To change the verification schedule:
1.
In the Administration view, click Database > Admin > Table Check.
2.
Using the Table Check Schedule pop-up lists, select the time at which you want verification to occur and the days of the week upon which you want it to occur.
3.
At any time during this process, click the Save Settings button to lock in your selections.
4.
If you would like any errors in tables automatically repaired, check the Enable automatic repair of corrupt tables box; otherwise, database errors will be logged but not repaired. If you’re having tables repaired, select the Repair Type using the pop-up list. The values, derived directly from the MySQL REPAIR TABLE command or myisamchk binary, are listed here in order from the least-invasive repair to the most-invasive. Since it is not always clear which level of repair to use, it makes the most sense to try the least invasive first and leave the most-invasive, USE_FRM, as a final resort. The levels include:
Quick
A Quick repair only tries to repair the index tree, not the data. (Equivalent of myisamchk --recover --quick.)
Normal
A Normal repair checks data for errors, repairs errors in both the index file and the data file, then rebuilds all indexes. It can be a slow process if you include the FullText Index. (Equivalent of myisamchk.)
Medium-Check
Scans rows to verify that deleted links are valid. (Equivalent of myisamchk --medium-check)
Extended (Slow)
An Extended repair, because of its slowness, should probably only be tried after you have already tried the others without successfully rooting out the problem. It does a full key lookup for all keys for every row. (Equivalent of myisamchk --safe-recover)
USE_FRM
USE_FRM is a complete table index rebuild from the data, superseding any of the other repair levels. Use it as a last and final resort.
[optional] To save some time, consider using the Skip Unchanged Table option.
5.
[optional] Click on the Check Now button to begin checking and/or repairing tables right away, rather than waiting for the scheduled time.
The View Log button provides a shortcut to the Logging, Table Check page where you can see the results of your most recent Table Check.