Great Circle Associates Majordomo-Users
(April 1996)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: It wasn't meant to be.
From: World of Coasters <jim @ rollercoaster . com>
Date: Mon, 29 Apr 1996 13:06:22 -0700
To: majordomo-users @ greatcircle . com

OK, I installed Perl 5.002, re-installed majordomo,
set all the permissions, and maajordomo test showed the
correct gid and uid.

So, I create a group and then mail a request 'subscribe coasters'
to majordomo. After another 15 minutes of heave load usage,
I get a response, 'out of memory'. I reboot Linux and funny, majordomo
starts up again and mails back another response:

>>>subscribe coasters

no confirmation, and the coasters file has no entries.

So, I reboot again, kill my httpd process. So now, nothing
but the basics are running. top shows 3400K of free mem
and the swap file is still at 30000K. I mail another
request, subscribe coasters. and then run top on my
console. The same thing happens. First, it gobbles up the 3000K
in mem then it slowly shrinks the swap file. It stays this
way for about 10 minutes then it finishes and I *still*
get the dreaded out of memory error!

Here's what my top looked like at the peak mem usage:

 11:52pm  up 12 min,  1 user,  load average: 2.87, 2.33, 1.253
27 processes: 26 sleeping, 1 running, 0 zombie, 0 stopped3
CPU states:  1.1% user,  5.2% system,  0.0% nice, 93.9% idle3
Mem:   6984K av,  6856K used,   128K free,   568K shrd,   404K buff3
Swap: 30740K av, 30684K used,    56K free3
3
  PID USER     PRI  NI SIZE  RES SHRD STAT %CPU %MEM  TIME COMMAND3
  151 bin       28   0 32003 4752  820 D     2.4 68.0  0:57 perl /usr/local/maj
  161 root      17   0   98  192  296 R     2.3  2.7  0:12 top3
  197 root       4   0  100  144  276 D     0.8  2.0  0:00 csh -f /usr/local/bi3
   73 root       6   0   62   24  200 S     0.5  0.3  0:02 in.telnetd3
   34 root       3   0   49   56  212 S     0.1  0.8  0:01 /usr/sbin/crond -l83
    1 root       1   0   56    0  184 SW    0.0  0.0  0:00 (init)3
  107 root       1   0   48    0  184 SW    0.0  0.0  0:00 (agetty)3
   47 root       1   0   53    0  168 SW    0.0  0.0  0:00 (syslogd)3
    6 root       1   0   27   72  256 S     0.0  1.0  0:00 (update)3
    7 root       1   0   27   72  256 S     0.0  1.0  0:00 (update)3
   49 root       1   0   40    0  172 SW    0.0  0.0  0:00 (klogd)3
   51 bin        1   0   60    0  188 SW    0.0  0.0  0:00 (rpc.portmap)3
   53 root       1   0   62    0  180 SW    0.0  0.0  0:00 (inetd)3
   55 root       1   0   31    0  180 SW    0.0  0.0  0:00 (lpd)3
   60 root       1   0   84    0  208 SW    0.0  0.0  0:00 (rpc.nfsd)3
   58 root       1   0   75    0  200 SW    0.0  0.0  0:00 (rpc.mountd)3
   74 root       1   0  107   60  236 S     0.0  0.8  0:00 (sendmail)3
  154 root       1   0  115    0  184 SW    0.0  0.0  0:00 (sendmail)3
  150 root       1   0  119   60  236 S     0.0  0.8  0:00 (sendmail)3
   80 root       1   0  168    0  192 SW    0.0  0.0  0:01 (tcsh)3

I am starting to think that majordomo does not like 8Megs. Can someone
please tell me how long majordomo takes to process a request on your
machine? 10 minutes seems *way* too long.

Jim
--
Jim Serio jim@rollercoster.com
Maintainer: World of Coasters - http://www.rollercoaster.com
The Web's Premier Coaster Site!




Follow-Ups:
Indexed By Date Previous: Re: Problems getting it going..
From: Keith Reding <kreding@info.aphis.usda.gov>
Next: Re: Archive problem
From: Michael Satterwhite <satterwh@insync.net>
Indexed By Thread Previous: Re: Archive problem
From: Crompton Administrator <admin@marconi.crompton.com>
Next: Re: It wasn't meant to be.
From: Keith Reding <kreding@info.aphis.usda.gov>

Google
 
Search Internet Search www.greatcircle.com