Steve K
posted this on January 11, 2011 15:15
When you get a *data corruption* there is no systematic, orderly, explainable way in which it happens. It is a random occurrence, which happens for various reasons such as:
Data corruptions happen randomly and it is impossible to pinpoint the exact cause. In order to prevent it from happening in future, all of the above has to be checked.
A great deal of cost and inconvenience can be avoided if you run the Data Integrity Check (Verify only) every day. This is usually done by the supervisor and by going to File -> Data Integrity and select verify only. This will then give you any errors that could be occurring on the system.
You should also make backups every day (or how much days data you're willing to have to re-enter should something go wrong), so that you can restore the backup when you get an error. This is probably the most important point to avoid losing the data.
You must also make sure that you are not doing date-out-of-period processing as this can also cause data corruptions.