procmail
[Top] [All Lists]

Re: NFS screwup -> hung procmail

1997-12-19 06:31:41
"PG" == Philip Guenther <guenther(_at_)gac(_dot_)edu> writes:

PG> Keith Pyle <keith(_at_)ibmoto(_dot_)com> writes: ...
There seems to be a bug where locks between the client (running AIX) and
the fileserver (Auspex) get lost.  To completely clear the problem, we
have to:
PG> ...
I would like to have an alarm timer around the procmail locking calls so
that the lock attempt would only block for a limited time.  If the lock
attempt fails, the mail could be requeued, and (even better) log a
warning.  Does anyone know if something like this is possible with the
current code?

PG> There's currently no way to do that, and I would not suggest adding it.
PG> If it's just a matter of kernel locking not working between the NFS
PG> server and client, recompile procmail without kernel locking (look for
PG> NO_fcntl_LOCK and friends in config.h) and run with just dotlocking.
PG> Then complain to your vendors again, suggesting that their failure to
PG> implement locking correctly means you'll have to buy your next server
PG> from someone else, say, Sun.

You can also write a script that kills procmail processes that have been
around too long.  We are doing this on one of our systems until we get all
of the kinks worked out.  The mail gets requeued.

-- 
Mark Plaksin                                http://www.arches.uga.edu/~happy/

<Prev in Thread] Current Thread [Next in Thread>