Date: Wed, 21 Feb 1996 12:20:04 -0600 (CST) From: "Mike Pritchard" <mpp@mpp.minn.net> To: bugs@freebsd.netcom.com (Mark Hittinger) Cc: hackers@FreeBSD.org Subject: Re: processes wouldn't die Message-ID: <199602211820.MAA03068@mpp.minn.net> In-Reply-To: <199602211447.IAA12886@freebsd.netcom.com> from "Mark Hittinger" at Feb 21, 96 08:47:24 am
next in thread | previous in thread | raw e-mail | index | archive | help
Mark Hittinger wrote: > > > > From: "Philippe Charnier" <charnier@lirmm.fr> > > > > Since some weeks, I run xsystats+rpc.rstatd to get information. rpc.rstatd > > receive signal 11 one or two times in a day. I also got the following > > configuration (reported by top) where I was not able to run the ps > > program nor to kill ps and rpc.rstatd. > > 11130 root -18 0 484K 12K lockrd 0:00 0.00% 0.00% ps > > 11076 charnier -18 0 480K 12K lockrd 0:00 0.00% 0.00% ps > > 11088 charnier -18 0 480K 12K lockrd 0:00 0.00% 0.00% ps > > 11100 charnier -18 0 480K 12K lockrd 0:00 0.00% 0.00% ps > > 11143 root -18 0 476K 12K lockrd 0:00 0.00% 0.00% ps > > 11154 root -18 0 476K 12K lockrd 0:00 0.00% 0.00% ps > > 10967 root -18 0 208K 12K lockrd 0:00 0.00% 0.00% rpc.rstatd > > I'm seeing this problem also. I believe that the 'ps' processes are hung > trying to read something out of swap space. It feels like a deadlock > condition to me. > > Philippe, could you tell us about your hardware configuration and what sort > of software packages you run on that box? > > For the 2.1 historically conscious I had these problems for awhile in the > sups after the 2.0-SNAP-0211 cut. After a few weeks of sup'ing the problem > went away. It returned a few weeks prior to the 2.1R cut and has been there > ever since. Are you running "pppd" on your machine? I just tracked down a the same type of hang on my machine, and determined that something in the kernel mode PPP was trashing the swap list somehow. If I disabled BSD style compression (the -bsdcomp flag), everything was fine. So far I've been unable to figure out exactly what in the bsd compress code is causing the problem. This may also be related to the problems other people are seeing with "top" and "swapinfo" being unable to read the swap list, since that is also a symptom of the problem. -- Mike Pritchard mpp@minn.net "Go that way. Really fast. If something gets in your way, turn"
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199602211820.MAA03068>