Journals

["Jeffrey M. Carey" <jeffreycarey@xxxxxxx>]

Has anyone else noticed what we think might be a major bug in the way
the AS/400 applies journal changes?

If a reorg. is ATTEMPTED on a file, the journal records it.  When
applying journal changes, the process stops when it reaches a reorg in
the journal - it can't reorg because the journal doesn't know if you
used a key file (it doesn't retain the parameters).  So you have to
mannually reorg. the file and start at the journal entry after the
reorg.

We were applying journal changes to recover after a job corrupted data.
We hit a RG in the journal, so we manually reorg'ed the file.  THe apply
failed on the next change to this file, saaying we were trying to change
a record that didn't exist.

As it turns out, the reorg. that the journal was recording FAILED.  Not
note in the journal that this happened.  So the manual reorg we did
during the apply left only twop records, so the journal couldn't apply a
change to rel. record 30.

More scary, what if that file had more records:
RGZPFM is submitted - "RG" entry in journal
RGZPFM fails    - no journal entry
record 30 is changed - "UP" entry in journal

APYJRNCHG starts 
APYJRNCHG hits "RG" and stops
user manually reorg. file
APYJRNCHG starts at next entry
APYJRNCHG changes the data in record 30

Problem - it's a different record 30!

Anyone else noticed/dealing with this?
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* This is the Midrange System Mailing List!  To submit a new message,   *
* send your mail to "MIDRANGE-L@midrange.com".  To unsubscribe from     *
* this list send email to MAJORDOMO@midrange.com and specify            *
* 'unsubscribe MIDRANGE-L' in the body of your message.  Questions      *
* should be directed to the list owner / operator: david@midrange.com   *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *