The suggestion made by Ray using uvbackup as a check of files is the
philosophy we have adopted after experiencing this type of corruption. We
run this every Saturday night over all our files and email the result
reporting any likely problems.
We are still discussing this one with IBM. They
I haven't really been following this thread, but have you tried the old trick of
performing a full backup of the account using uvbackup, redirecting stdout to
/dev/null (or .\NUL for those on Windows) and redirecting stderr to a file in which
any corrupted files will be reported?
The nice thing