My problem was:
>
>For the second time we have got a "cat" process hanging which can't be killed
>using kill -9:
>
>chris 2550 0.0 0.0 124 0 p4 IW 11:01 0:01 -csh (csh)
>chris 10738 0.0 0.0 12 0 p4 DW 17:20 0:00 cat N50P200sig1phi.log
>
>Process 10738 hangs.
>The system is a Sun 4/75 running SunOS 4.1.1, a diskless client.
>The kernel is the sun-supplied DL60. We applied patch 100338-02: SunOS 4.1;
>4.1.1 system crashes with assertion failed panic.
>
>The /dev/ttyp4 file shows that can be used, but when login tries to access
>p4 the user is logged out right away without any messages.
>
>The problem then is that no more then 4 users can be logged on before rebooting.
>Any ideas on how to kill the "cat" process?
>Is it a NFS bug?
>
Most people pointed out that the process cant be killed because it is in a
non-interruptible wait. That is nice, but (a) why does it hang, and (b) why
is
/dev/ttyp4 (semi) accessible for other users, while it is used by Chris?
To solve problem (a) several people suggested to install the NFS jumbo
patch
(100173-03). We applied the patch a week ago, and don't have any problems
ever since.
No one made any suggestion about problem (b): it remains a mistery to me.
Many thanks to:
Jason Ornstein <ornstein@refuge.Colorado.EDU>
mark@isis.Berkeley.EDU (mark kraitchman)
kalli!kevin@fourx.Aus.Sun.COM (Kevin Sheehan {Consulting Poster Child})
ray@isor.vuw.ac.nz
kwthomas@nsslsun.nssl.uoknor.edu (Kevin W. Thomas)
Ace Stewart <jstewart@mailbox.syr.edu>
dal@gcm.com (Dan Lorenzini)
Mike Raffety <miker@sbcoc.com>
jem@matt.ksu.ksu.edu (James Morrill)
stern@sunne.East.Sun.COM (Hal Stern - NE Area Tactical Engineering)
ross!hdr@cs.utexas.edu (Henry D. Reynolds)
celso@dcc.ufmg.br (Celso de Souza Lima)
Peter Barneveld Internet: peter@FenK.wau.nl
Wageningen Agricultural University Fax: (-)(31)8370-83777
Dpt. of Physical and Colloid Chemistry Voice: (-)(31)8370-82144
The Netherlands Feel free to ftp metten.fenk.wau.nl
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:22 CDT