It's a possibility. Could you verify that the system is swapping, not
just discarding cached pages and re-reading them, by running 'vmstat 1'
while the system is in this state (and stopping top at some point in the
trace)?
> I'm using 2.2.10 kernel with a "restricted proc" patch. But this
> problem is stands from the early 2.2 until 2.2.10 (>2.2.10 I had
> problems, that's why I use that).
> I have tried 2.3.17 also, and it seems to have the similar
> problem, but I have some strange IPC problems, so I didn't test it.
Could you elaborate on the problems you're having so they can get fixed
allowing you to test 2.3.recent? 2.3.16+ have Andrea's LRU page cache
patch in, and as such have a much better replacement strategy for cached
data which should eliminate the behaviour you're seeing, so it would be a
very useful comparison.
-ben
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/