[error] Apache2::RequestIO::read: (70007) The timeout specif

[jamiegau]

Recently, my OTRS install developed the following error..
[error] Apache2::RequestIO::read: (70007) The timeout specified has expired at (eval 112) line 5

When I try to respond to a ticket.
It started out of the blue.. It may have started after a apt-get update..
My platform is ubuntu, was 10.04 LTS
I decided to upgrade to latest everything to see if it would help.
So now on ubuntu 12.04 LTS (Which I highly recommend in general) and OTRS 3.1.5
The upgrade went OK.. but the problem still remains..

Could anyone give me a hint on what to look for? Is this come type of incompatibility with upgraded perl or perl modules? or Apache?
Google it has not helped.

I am surprised no one else has hit this problem!!

Thanks,
James

[CSL]

If this only happens when you try to respond to a ticket, then system could be having problems sending mail. Have you checked your email settings? Have you looked at the apache error log as well as OTRS's log?

[jamiegau]

Thanks for trying to help, however, that does not seem to be the problem.

The system can send email directly no problem.
The OTRS logs show NO ERRORS.

The error I have mentioned is from
/var/log/apache2/error.log
[error] Apache2::RequestIO::read: (70007) The timeout specified has expired at (eval 112) line 5

So. Trying to check what you suggest has shown no light on this.

If anything, it is strange that the OTRS log shows NO ERRORS. Only the apache log shows the error.

Do you have any other suggestions on what to look for.
Or how I can get more error logs on the possible problem...

Thanks,
James

[CSL]

Are you using mod_perl, FastCGI, or CGI in Apache? It would be worth switching between mod_perl and FastCGI to see if both of them produce this error.

Information on this error is scarce as you say; I've only found some references to possible mod_perl issues and some proxy timeout problems. I take it you're not using a proxy (forward or reverse) with this machine?

[jamiegau]

Ok,
This problem is getting even wired-er...

Get this.. From HOME to my OTRS server, the problem does not exist..

But from in the office.. Any system trying to use the OTRS server this problem occurs.
Now,
Office network->LinuxFireWall-running-Squid-transparent-Proxy->live-C-Class-with-OTRS-server

First thing I did was make sure the OTRS server could find a reverse lookup and placed my Firewall IP into the /etc/hosts file
Maybe some type of nslookup time out was causing an issue.. NO!!!!

So. this has me stumped. How could this setup cause systems internal behind my firewall/squidProxy to have this issue while those NOT behind the firewall no issue.

Note: the OTRS server is a VM on a system I run about 6 different servers.

From internal network I have tried different computers, deleted cache.. even tried on a MacPro as opposed to PCs we generally use. They seem to all cause this problem. But my home computer. works great and the error does not occur..
All I can think of is...
Could it be a Quid-Proxy issue?

James

[CSL]

So any machine you try that connects through this proxy server produced the error, while another machine that doesn't works fine with no errors?

I'd say it's good bet the proxy is the issue in that case. There could be many different reasons for this depending on what that proxy does. A transparent proxy may be changing the requests / answers in a way that OTRS's apache doesn't like for some reason. I've used a standard squid proxy with OTRS before with no issues, so it's probably something unique to your setup.

As for troubleshooting this, it depends on how knowledgeable / adventurous you are... comparing packet captures done on the OTRS box of both a working session and a session that produced the problem might be a place to start.