[Raw Msg Headers][Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: more disk spinnin' stuff



On Mon, 19 Mar 2001, Matti Aarnio wrote:

>   You don't see anything at  $LOGDIR/router  or  $LOGDIR/scheduler
>   files ?  (Anything peculiar which accumulates fast, that is..)

===== Smtp server log =====
000000000#	started server pid 17741 at Mon, 19 Mar 2001 17:45:53 +0000
000000000#	killed server pid 17741 at Mon, 19 Mar 2001 17:48:09 +0000
===== End =====

===== Router Log =====
router: router daemon (2.99.55 #1: Mon Mar 19 16:17:18 GMT 2001)
	started at Mon, 19 Mar 2001 17:46:03 +0000
===== End =====

===== Scheduler Log =====
scheduler: scheduler daemon (2.99.55 #1: Mon Mar 19 16:17:35 GMT 2001)
	pid 25208 started at Mon, 19 Mar 2001 17:46:21 +0000

Synchronous startup completed, messages: 0 (0 skipped) recipients: 0
***********************************************************************
scheduler: exit(0): signal
===== End =====



>   Could you pinpoint it to one or three subsystems:
> 		smtpserver
> 		router
> 		scheduler

Scheduler is the culprit. I built zmailer at work today (logs above)
and started each component individually like you said. Not until
scheduler started did the disk i/o start going to town.

Now, a bit more history... I have both ndbm and gdbm installed. gdbm
is in /usr/local/... and ndbm is in /usr/... While running the
configuration script, it finds ndbm, but not gdbm. If I "make" zmailer
like this, the make will puke out with an error, something to do with
ndbm. So, I copy gdbm.h to /usr/include, and the three gdbm libs to
/usr/lib, _and_ rename ndbm.h out of the way for the moment, it
configures and bulds okay.


--
Dana Booth <dana@booth-family.org>
Tacoma