We've experienced some strange behaviour with archive. Here's the scenario:

Client - ADSM version 2.1.0.2 for solaris running under solaris 2.4.
Server - ADSM/6000 version 2.1.0.2

Our mail directory on the solaris machine has over 22,000 mail inboxes in
it of the form /var/mail/userinboxfile.

I'm attempting an archive of a single file in the /var/mail directory using:

archive /var/mail/testfile

The archive client seems to lock up after archive processing is invoked.
The command works fine if we chose a file in a much smaller directory, such
as archive /tmp/testfile

We were curious what was going on, so ran truss to see what the archive
command was doing. Even though we specified a specific file in the
directory, the archive command was going through EVERY FILE in the
/var/mail directory doing an LSTAT on it to gather information. With 22,000
files this was going to take a very long time!

This behaviour makes it impossible to use archive on specific files in
directories that contain large numbers of files. The behaviour seems
strange to me - why do you need to lstat all 222,000 files in the directory
when only one was specifically selected for archive?

I seriously hope this is a bug that can be corrected or there is an option
that prevents this behaviour. Have others encountered this? Perhaps IBM can
comment.

P.S. Out of curiousity we also tried it with a version 1.2 client with the
same results.



Chris Krusch                             Email: Chris.Krusch AT ubc DOT ca
University Computing Services            Phone: (604)822-4215