I am slowly coming around to the idea of splitting my patch up into many parts, each addressing just one little issue, so there would be a couple of [PATCH]-es and some [RFC]-s, each profusely explained. Unfortunately did not yet have time to work on this; but want to do, hopefully soon.
However... we now did another test on the server, and it ran into OOM. We were load-testing. We intend to use the server for student logins, via XDMCP from any one of 109 X-terminals (similar to LTSP); we logged in on all our terminals, and on each ran a firefox and an rstudio. The server ran into OOM, having plenty of free memory but having exhausted lowmem, after 80 logins. Then rebooted the server with mem=32G on the kernel, and there was no OOM after logging on all our "fake" students. So, there is a bug still. I will try to make up some easily reproducible test, and if can provoke OOM then look again into kernel code. Cheers, Paul Paul Szabo p...@maths.usyd.edu.au http://www.maths.usyd.edu.au/u/psz/ School of Mathematics and Statistics University of Sydney Australia -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201301100148.r0a1m1pj022...@como.maths.usyd.edu.au
