Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Fork Error with CGI Programs on iPlanet 4.1

9 views
Skip to first unread message

Hans

unread,
Nov 15, 2000, 3:00:00 AM11/15/00
to
Hello

I am running iPlanet Enterprise Server 4.1 with SP3 on a Redhat Linux
6.1 server
with a PIII 500 MHZ processor and 128 MB of SDRAM. Since iPlanet needs
a 256 MB swap and I only had a 64 MB swap partition, I created a 192 MB
swap file.

I have several perl scripts (e.g. message boards) running on my server.
My server ran well for 21 days but after that, when I try to access
my message board (written in perl), I got a "HTTP 500 Internal Server
Error" randomly. Sometimes I would get the error and sometimes the
message board would work. When I checked the iPlanet server logs, I saw
this error message each time the message board failed:

[14/Nov/2000:17:46:29] catastrophe ( 1943): for host 216.123.4.5
trying to POST /bin/board.pl, cgieng_start_output reports: could not
fork new process (Request pipe I/O failed [Success])

Do I have some kind of resource problem on my server? Do I have enough
memory or a big enough swap file? Maybe the swap file I added is not
being utilized properly. Do I need to reconfigure the CGIStubs
directive? I may try rebooting to see if this clears the problem up for
another 21 says. Any help or advice anyone could give would be
appreciated. Thanks.

Hans
E-mail: ha...@futurecasting2000.com


Jimi Thompson

unread,
Nov 16, 2000, 3:00:00 AM11/16/00
to
That sounds like a serious resouce problem. Check your disk space. How
much is available?

Hans wrote:

--
Jimi Thompson
Web Master
L3 communications

"It's the same thing we do every night, Pinky."

jthompson.vcf

Hans

unread,
Nov 26, 2000, 3:00:00 AM11/26/00
to
Jimi

Rebooting my server got rid of all of the CGIStub processes that were running
and the fork error. Now I usually just see 4 CGIStub processes running when I
run netstat. I am still worried that I might have this CGIStub/fork error
problems again over time.

I read your response to my post on the newsgroup below but I still have a few
questions. I checked my disk space using the df -h command and I saw that my
resources are:

Filesystem Size Used Avail Use% Mounted on
/dev/sda1 197M 132M 55M 71% /
/dev/sda8 197M 15M 173M 8% /home
/dev/sda6 1.9G 848M 986M 46% /usr
/dev/sda7 387M 204M 164M 55% /var
/dev/sda5 5.4G 1.5G 3.6G 29% /www

the cat /proc/swaps command gives:

Filename Type Size Used Priority
/dev/sda9 partition 72256 6204 -1
/www/swapfile file 189880 0 -2

I looked in my magnus.conf file, but I didn't see any CGI directives
configured in there. Do I need to configure these CGI directives:

MinCGIStubs
MaxCGIStubs
CGIStubIdleTimeout

Can you give me any advice on how to configure the magnus.conf CGI directives
to optimize my server or do I just need more memory (I currently have 128
MB)? Or could the problem be with my swapfile? Thanks for the help.

Hans
-------

0 new messages