One thing I ran across in googling was that csh v 6.17 was actually a symbolic link to tcsh.

http://comments.gmane.org/gmane.linu...ase.rhel6/1525

which indicates that csh isn't handling exit codes properly (bug #784510). This could cause tcsh to process things it shouldn't.

csh probably somewhere has an unsleeping loop that waits for a condition or an event. You can try building and installing a new version of it (csh or tcsh) but make sure it doesn't break your system.

I am currently trying to find the error in the scripts. What I am doing is, I am printing timestamps into a log file which then gives me clues which script does eat up so much CPU (and taking 10 seconds for it).

**** code im inserting into various scripts
echo "`date` begin .cshrc" <</tmp/test.log

The output tells me that the following after having logged in at 15:22:25

Fri Aug 2 15:22:25 CEST 2013 begin etc/csh.login
Fri Aug 2 15:22:25 CEST 2013 end etc/csh.login
Fri Aug 2 15:22:25 CEST 2013 begin.cshrc
Fri Aug 2 15:22:25 CEST 2013 end.cshrc

-> this means that the script that has problems is executed after .cshrc

I have tried to google for "linux user login scripts" but had little luck, can anyone tell me which scripts are called after .cshrc??

There can be a .tcshrc... which includes .cshrc

It is also possible, though unlikely, that something is running in the background.

The .tcshrc file only includes the following:

[root@lin133 ~]# locate .tcshrc
/root/.tcshrc
[root@lin133 ~]# cat /root/.tcshrc
# .tcshrc

# User specific aliases and functions

alias rm 'rm -i'
alias cp 'cp -i'
alias mv 'mv -i'

set prompt='[%n@%m %c]# '

If it's not in the scripts then it could be csh itself like how I described it earlier.

Ok. Before I try to "reinstall" csh, do you know which scripts run after .cshrc (in the home directory) when a user logs in?

I don't know much about csh sorry. But I think it's well detailed in the Startup and Shutdown section of tcsh's manual.

So clearing the history file did it. Now the login works quick and cpu is at 1.7%.
I figure the history file was huge. I wonder why that is not the case on my other server that is still running centos 5.5
I have checked in bashrc in both servers and there is no comment that would limit the history file size.

Well good thing you found the solution already. Please mark the thread as solved.