On Jun 19, Ken Wellsch wrote
> I've been using this
> 
> 	options		NMBCLUSTERS=2048
> 
> for about a year now.  When the system started to get more use with
> multiple folks and web serving and such, I had to do this to deal
> with the routine variety of "mb_map full" messages... sigh.
> 
> I have watched via "netstat -m" as the percent used number grows
> rapidly when I'm running my application.  I mean rapidly too.  I
> ran "netstat" as quickly as I could and the %-used increased something
> like 6% to start, 23%, 56%, 73% and that is when I killed the process.
> 
> It only occurs between a Mac and our i386 based server.

Well, I have a problem with an I386 server too. It's a k6/200 with
64 Mb RAM, runnin 1.3.2 with apache 1.2.6. It has virtual web servers
(with interfaces adresses aliases). Once in a while (but several times in a
wek), the http server just hang. It's not overloaded (about 300-400
request/hour max).
It sits there with some datas in the incoming socket queues. a kill -1
on the master restart the thing (the incoming waiting connextions are not
killed, just waken up). Other network connextions are still alive (sendmail,
ftp, telnet, ...).
The weird thing is that this server has been very reliable for > 1 month
before starting behaving this way.
The other weird thing is that I have another server running 1.3_BETA, same
apache 1.2.6, with about the same load. Works like a charm. It's a 486/33
with 8 Mb RAM ...

For now I didn't have the time to track this down. I've collected a core
dump (reboot -d -q) of the system while httpd was hung. I didn't see anything
special (only 10% use in netstat -m. It has free kernel memory too).
This machine is running with
options NMBCLUSTERS=2048
options NBUF=2500
options BUFPAGES=2500
options NKMEMCLUSTERS=2048

I'm not sure at all the 2 problems are related, but they may be.

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--