(Illustration by Gaich Muramatsu)
On Thu, 30 Aug 2001 13:21:18 -0400, Jan Harkes said: >On Mon, Aug 27, 2001 at 08:19:57PM +0200, Florian Schaefer wrote: > > I attached the output after I did "cfs cs" because I didn't want to > > clobber the output with the linebreaks of this mail. > > Ok, I've taken some time to look at this trace. Thanks a lot for taking the time. > The server crash is not triggered by 'cfs cs', but the reintegration > that starts when the servers are found to be up. The crash happens at > the end of the reintegration phase, strangely enough while dereferencing > a pointer that was allocated earlier on in the reintegration. I don't > see how that pointer can get lost/corrupted. > > The simplest solution for now is to purge the modification log on the > client so that it won't try to reintegrate the local changes that are > triggering the crash. What is the modification log? I created the tarball backed it up and deleted the wohle dir witch my userid. Furthermore I moved /usr/coda/LOG to another place but as venus would not start then I decided that this file is important and put it back to it's orinignal place. So I ended having deleted /usr/coda/spool/<userid>. But "cfs lv" still shows the pending entries, venus still tries to sync the changes with the server, oh and of course, crashes vice. What have I done wrong? Ciao FlorianReceived on 2001-08-31 01:28:19