I’m in the process of rolling out archiving to my organisation.
While in the process of testing, when I first specified the setting and ran compact -A against only the one mail file, I got the following result in the server’s log:
26/11/2007 15:55:52 Archived mail/user.nsf, 8499 documents were archived and 8374 were deleted
When I ran compact against the whole directory the next day, I got this:
27/11/2007 12:23:23 Archived mail/user.nsf, 372 documents were archived and 460 were deleted
I don’t think my archive selection formula is strange - Server to Server, Archive anything not modified in last 90 days, Copy and then clean up this database, Don’t delete documents that have responses.
Any ideas why I’m getting such random results? I’m not as worried about the former - extra documents aren’t a biggie, but mail vanishing could be an issue.
It’s not just one off random behaviour. With another one I got Archived 3225, delted 3185, but when I actually open up the archive I get a document count of 3258.
Then 8499 were archived and 8374 deleted, by a compact -a.
The following day, 372 documents were archived and 460 were deleted, also by a compact -a.
The original database now has 162 document count.
The archive now has 8958 document count.
It’s a normal 6.5.5 mail file, with some very minor modifications (Reply with History is modified to not contain attachments, room reservations trigger an email to administrators) but nothing that should be affecting archiving.
I have some more databases I can test with. Do you have any suggestions of what stats to take before and after, screenshots, etc, to try to work out what is going on?
It could be parent and child weirdness - but I still wouldn’t expect the deleted count to be higher than the archive. Orphans shouldn’t delete just because they’re orphans should then?
Maybe it’s calendar stuff - repeating meetings or something.
You say there were more documents deleted than archived, but could they be ones archived previously that were not deleted because the parent was still there? The only figure that looks wrong to me is the count in the Archive, which seems high. unfortunately, i can’t think of a reason just at the moment, sorry.
I ran another test, this time taking a local copy of the mailfile at each stage of archiving. I created a private view, with no hierarchy, showing me the docunid and subject.
Again, 9 extra documents were deleted in the second wave of archiving, but as I had the before and after snapshot, I could compare what was missing.
The deleted documents had all already been copied to the archive BUT didn’t show up in the mail thread view and were, therefore orphans.
Thanks so much for your help. I think I can roll out archiving the masses with some more confidence that nothing will vanish.
If I find the cause of the too high doc count, I’ll post the reason.