[rhn-users] RHEL3 box hanging at login
Tom Hodder
tom at ecnow.co.uk
Sun Nov 13 13:17:02 UTC 2005
Hi,
In a blade farm of a hundred or so RHEL3 boxes, 2 of these machines started
hanging after the password was entered at the login prompt, both via ssh and
locally, this was after a few days of running time.
This didn't seem to be effecting services such as httpd, but atd and crond were
also hung. I guess atd and crond try to change to a user context and require
login functionality to do that.
I've been troubleshooting the problem this weekend, so I stopped some services
to get a smaller set of processes that might be causing this problem and
rebooted.
BigMistake!(tm) - Now the box is hanging at the first prompt, though services
like apache seem to have started up.
I understand that there is a known bug/condition that is caused by ssh sessions
ending and not killing their children, and this seems to be what is happening
here. Though I don't understand why it effects the local logins as well.
Has anyone seen this problem before and got any ideas on what might be a good
line to take in troubleshooting this machine.
Tom
More information about the rhn-users
mailing list