home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.linux
- Path: sparky!uunet!gatech!darwin.sura.net!zaphod.mps.ohio-state.edu!moe.ksu.ksu.edu!engr.uark.edu!cseg03.uark.edu!dws
- From: dws@cseg03.uark.edu (David W. Summers)
- Subject: Re: What causes these Zombie processes?
- Message-ID: <1993Jan10.205719.1135@cseg03.uark.edu>
- Sender: netnews@engr.uark.edu (NetNews Administrator)
- Nntp-Posting-Host: cseg03.uark.edu
- Organization: University of Arkansas
- References: <1798@emoryu1.cc.emory.edu>
- Date: Sun, 10 Jan 1993 20:57:19 GMT
- Lines: 52
-
-
- In comp.os.linux you write:
-
- >After running for a few days, my Linux system ends up with quite a
- >few 'Zombie' processes. I can't kill them off, even with kill -9 and
- >the only way to get rid of them is with a reboot. This is a typical
- >'ps -aux' screen after a few days.
- >
- >USER PID %CPU %MEM SIZE RSS TT STAT START TIME COMMAND
- >root 1 0.0 3.3 32 136 ? S 23:39 0:00 init
- >root 600 0.2 10.7 300 440 a1 S 10:06 0:07 -sh (sh)
- >root 3 0.1 0.9 2 40 ? S 23:39 0:55 /etc/update
- >root 376 0.0 2.8 48 116 a2 S 09:30 0:00 /etc/getty 9600 tty2
- >root 322 0.0 0.0 48 0 a3 SW 08:45 0:00 (getty)
- >root 323 0.0 0.0 48 0 a4 SW 08:45 0:00 (getty)
- >root 22 0.0 2.9 80 120 ? S 23:39 0:00 /usr/etc/inet/portmap
- >root 24 0.1 4.3 56 180 ? S 23:39 0:50 /usr/etc/inet/inetd
- >lee 26 0.0 0.0 112 0 ? SW 23:39 0:29 (nfsd)
- >root 28 0.0 3.4 80 140 ? S 23:39 0:00 /usr/etc/inet/mountd
- >root 318 0.0 0.0 96 0 ? Z 08:21 0:00 <defunct>
- >root 319 0.0 0.0 96 0 ? Z 08:44 0:00 <defunct>
- >root 341 0.0 0.0 96 0 ? Z 08:58 0:00 <defunct>
- >root 622 0.0 8.1 268 332 a1 S 10:10 0:01 sh /usr/bin/X11/startx
- >root 463 0.0 0.0 96 0 ? Z 09:41 0:00 <defunct>
- >
- >Notice that the processes are all 96k in size. The process number still
- >shows up in the /proc partition, but I don't know how to get any info
- >from that. I'm open to suggestions and advice.
- >
- >I'm running mostly SLS 0.98p5 with a 0.99p2 kernel, netbin0.1, X11, and
- >smtp smail. The system also mounts and serves up NFS partitions.
- >
- >Thanks
- >
- >--
- >Seng-Poh Lee <splee@pd.org>
-
- The problem is probably the 'inetd' daemon. I was having the same problem
- and found out that everytime after I did a 'finger' of my Linux machine
- that another <defunct> process would appear. I put debugging code in
- the 'inetd.c' file and found out that the first time the SIGCHLD signal
- was being sent the inetd program was catching it but none thereafter.
-
- I put a 'signal( SIGCHLD, reapchild);' at the end of the reapchild
- subroutine and that fixed it.
-
- - David Summers
-
- --
- "Never under-estimate the bandwidth of a station-wagon
- David Summers full of tapes, hurtling down the highway."
- dws@engr.uark.edu - Tanenbaum, "Computer Networks"
-