we changed the time out in httpd.conf:

# Timeout: The number of seconds before receives and sends time out

Timeout 1800

On the other hand, how may processes do you allow in
# MaxKeepAliveRequests: The maximum number of requests to allow
# during a persistent connection. Set to 0 to allow an unlimited amount.
# We reccomend you leave this number high, for maximum performance.

MaxKeepAliveRequests 100

Probably this will help.

Best regards, Andreas Schmidt, CSC, Germany





Peter van Boheemen <[log in to unmask]>@RAVEN.UTC.EDU> on
26/03/2001 04:30:35 PM

Please respond to Peter van Boheemen <[log in to unmask]>

Sent by:  HP-3000 Systems Discussion <[log in to unmask]>


To:   [log in to unmask]
cc:
Subject:  [HP3000-L] Apache web server sometimes won't spawn new child


We have been using the HP3000 as a web server since 1995. We started with
the NCSA server, then moved to the OMI web server. Recently we have moved
to the Apache web server. Since we did that we experience some problems.
Most of the request to our web server will result in the execution of our
WebQuery cgi program, which will perform our database queries and updates.
A few times a day, probably when it is busy, the web server returns
an 'Internal server error'. In the log file we see the reason for
this:"Resource busy, try again: couldn't spawn child
process: /WWW/WWW/httpd_1.3/cgi-bin/WebQuery"

It looks like Apache invokes a time out, when then HP does not spawn a cgi
process fast enough. We never experienced this problem with the OMI server.
Has anyone ever had this experience ?
Any suggestions ?
regards,
Peter van Boheemen
Wageningen University Library
The Netherlands