Follow-up Comment #7, sr #105931 (project administration):
This is not a user-side problem. It's happening for me too; it's hard to get
any work done. I can detect no pattern to when it happens and when it
doesn't.
It seems apparent that the 500mb claimed to be available in /tmp isn't really
available. Can we make /tmp just be on disk? (Or point cvs to some other
directory; I'm not sure if that can be done without recompiling.) I know
it's slower, but right now we're having a much worse problem. I would bet
this problem is related to tmpfs. Anyway, 1gb is not enough.
As for usage, there are around 2400 cvs-serv directories. None are very big
in themselves, but that adds up to about the right amount. They go back to
June 20. Can we delete some of the older ones? At least then the immediate
breakage should be relieved.
Karl
_______________________________________________________
Reply to this item at:
<http://savannah.gnu.org/support/?105931>
_______________________________________________
Message sent via/by Savannah
http://savannah.gnu.org/
