[svlug] sendmail NOOP slowdown

Rick Moen rick at linuxmafia.com
Mon Feb 11 11:34:50 PST 2013

Quoting Aleksandr Miroslav (alexmiroslav at gmail.com):

> What is a problem right now is the code that is actively running. We
> need to flush out everything we have.
> Is there anyway to tune sendmail to allow these to go through right
> now? The only option I see is to kill -hup the sendmail child every 2
> minutes or so, the client code then restarts a new connection and
> sends out a bunch more emails before being tarpitted again (we have
> about 4k emails left to send out), this would flush the queue in a
> couple of hours.
> But that's a very ugly solution. I wonder if there is something
> cleaner we can do.

That's actually how *I'd* do it.  Sure it's hack'n'slash system
administration, but it quickly gets you past a very temporary problem
and gets the job done.  The search for elegant and optimal solutions is
not necessarily your friend, especially for problems that aren't going
to persist very long.  

As an aside, you _may_ find that expertise specifically with sendmail is
getting rarer, these days.  Not that I would ever argue with a fellow
sysadmin's choice in MTA religion.  ;->

More information about the svlug mailing list