Hi All

Had an issue recently with our eDir (IDVault) to eDir (SSO )driver
whereby a server in the connected SSO tree crashed and took some time
to get back up again.

Initially this didn't cause us any issues until somebody moved two
users between OUs, which the downed server was in the replica ring for.
This caused the second user move operation on the driver to fail and
generate a -637 Previous Move In Progress.

This meant that all subsequent eDir to eDir events, even ones not
related to the problem OUs were then stuck in the queue waiting for the
first events to clear.

I did think about using dxcmd to remove events, but in the end I just
forced the downed problem server out of the replica ring, ran a repair
to clear up the backlink to non existent server, then the obits all
processed automatically and everything lept back into life.

Does anyone have any suggestions on how to deal with this type of
problem in the future in as clean a way as possible ?


--
clovercne
------------------------------------------------------------------------
clovercne's Profile: http://forums.novell.com/member.php?userid=3215
View this thread: http://forums.novell.com/showthread.php?t=336157