On Mon, May 29, 2006 at 02:09:29PM +0200, Longina Przybyszewska wrote:

I have a group of clients which recently I started to reboot daily. Now, they never recover from the purple on cpu/disk/memory.. etc monitoring. Client processes run fine, but in clientdata.log, there are a lot of: --- 2006-05-29 13:32:43 hobbitd_client: Got message 12359 @@client#12359|1148902363.482265|130.225.128.99|leonora.imada .sdu.dk|linux

2006-05-29 13:32:43 2006-05-29 13:32:43 Worker process died with exit code 134, terminating

This means hobbitd_client which parses the client message has crashed. Please try upgrading to the latest snapshot. If that doesn't work, send me a copy of the client message from that host (you'll find it in the ~hobbit/client/tmp/msg*.txt file), and please run any core-dumps in ~hobbit/server/tmp/ through gdb to get a stacktrace - see http://www.hswn.dk/hobbit/help/known-issues.html#bugreport Regards, Henrik