httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jonne Zutt <j.z...@ewi.tudelft.nl>
Subject [users@httpd] clients get stuck loading for ever
Date Tue, 02 Mar 2004 10:29:09 GMT
Dear apache users,

If a webserver functions perfectly locally (always), and sometimes works
properly remotely but after some time gets broken by some mysterious
reason, how can one diagnose what is going on?

It seems to have nothing to do with our firewall (the problem remains
without it and it is well configured).

I get no error messages anywhere (I use httd.conf's debug level
already). The problem is that clients get stuck loading pages for ever
(pages that worked before, file permissions are okay).

I figured out that if a client gets stuck, netstat displays the
following:
tcp 0 10890 130.161.159.32:80  130.161.158.125:51210  FIN_WAIT1
which closes and then soon disappears if I close this hanging client's
browser:
tcp 1 10890 130.161.159.32:80  130.161.158.125:51210  CLOSING
I don't know what can be deduced from this information.

I don't know for sure it's a httpd problem, it might be something else.
I have a working network connection though. (so what else can it be i'm
thinking)

Is there some caching going terribly wrong somewhere? How can I figure
that out? I already saw caching was disable in my (and the default)
httpd.conf file.
Even rebooting my machine didn't help me this time to get my site to
work again.

Does anyone have any debugging tips?
Thanks in advance,
Jonne.

ps, I'm running:
- fedora core 1
- httpd 2.0.48
- 3c2000 network driver
- I attached my httpd.conf, I hope you don't mind...

Mime
View raw message