Original question :
> I have a strange problem with a comsat server process on IPX running 4.1.3.
> It affects only one user who logins from a Mac. His .cshrc and .login
> files do not differ from other people files, nor other Mac users have the
> same problem. The problem is that in.comsat never dies for that user, and
> I have to manually kill it after a while. The ps reads the folowing:
> user 26134 0.0 0.0 36 0 ? IW 20:58 0:00 in.comsat 240
> user 26840 0.0 0.0 36 0 ? IW 05:43 0:00 in.comsat 240
> user 27031 0.0 0.0 36 0 ? IW 08:43 0:00 in.comsat 240
> user 25980 0.0 0.0 36 0 ? IW 18:36 0:00 in.comsat 240
> user 26553 0.0 0.0 36 0 ? IW 03:27 0:00 in.comsat 240
> user 26620 0.0 0.0 36 0 ? IW 04:43 0:00 in.comsat 240
>
Thanks for their responses to
craven@cg.NRCan.gc.ca
thielen@irus.rri.uwo.ca
jim@nga.com
In my case, unfortunately, none applied. I have searched a bit more and
I hope I have the answer. The problem is not on the SUN side, but on a Mac,
in the MacX settings. The user have changed (and forgotten about this) the
path, and was using /usr/openwin/bin/xterm, not an xterm from X11R5.
I put a proper path to an xterm and all the comsats are properly exiting.
So far one week passed by, and I had no comsats :)
Nadya Williams, UCSD
-----------------------------
>From craven@cg.NRCan.gc.ca
>We had much the same problem. The source of our problem was some bad
>entries in the utmp file which were causing the comsat daemon to fire
>up whenever the user associated with the bad entries logged in and
>issued a 'biff y' command. See the man page on in.comsat for a little
>more info. Removal of the bad entries in the utmp file (in our case
>by I think we killed a few csh processes) will likely rid you of the
>extra comsat's that don't die. Our OS was sunos4.1.3_u1 so if yours is
>solaris the above may not apply.
>
>Good luck,
>
> Jim
no bad entries in utmp on my system
-----------------------------
>From thielen@irus.rri.uwo.ca
>
>Hi!
>
>There is a patch for this...
>
>check out http://sunsite.unc.edu/pub/patches/patches.html
>
>I had to apply one for the same sort of thing.
>
>sue
I have installed that patch (Patch-ID# 100272-07) in 1995...
---------------------------------------
> From jim@nga.com
> Hope you get an answer. Maybe it will narrow the search a bit
> to know that I had the same sort of problem on a sun ipc 4.1.3.
>
> Used to run Biff on the console window. Things would be fine
> for a week or so then one day there would be 200 comsats......
>
> good luck
> Jim Roy
just like I had, except things were fine for a day or so...
-----------------------------------------
> From thielen@irus.rri.uwo.ca
> If the user is on a machine that / fills up on, and has biff
> running, you'll also get a load of these comsats hanging around
> until / gets space and biff can print out to the console..
>
> sue
no sign of / filling up...
----------------------------------------
> From thielen@irus.rri.uwo.ca
> Comsat isn't dieing because biff isn't getting it's output
> out. See whether or not biffs output is working or not...
>
> I think...
>
> THen get them to not use it until you find a fix
>
> sue
the biff was set to "y" and working.
---------------------------------------------
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:11:44 CDT