tynor@prism.gatech.EDU (Steve Tynor) (12/12/90)
I've been having bad luck lately in that I seem to keep making my Apollo crash.
I always happens shortly after I start a 'make' in an xterm. I suspect that the
machine is running out of process table, but that's just a guess. This is
happening a couple times a day!
None of my colleagues have ever had this problem, so I suspect that it has
something to do with the software I run: DN3500, 8M, OS 10.2, twm 1.2, 2 or
three xterms running berkely sh and bash 1.05, gnuemacs 18.55 (with Leonard
Zubkoff's Apollo patches - using the X version, not the pad version), xclock,
xbiff, xload. I'm the only person in the group using bash, emacs, and an X
window manager.
When I called 1-800-4APOLLO, they suggested reloading the OS. That seemed to
fix it - I've gone a couple of months without a single problem. Now I'm using a
different machine (a DN4500, 16M), and am seeing the same sort of behavior.
Has anyone else had this sort of problem? Is it indeed related to the software
I'm using? How is it possible that user-mode programs can corrupt the OS???
Is there some way that I could configure the machines to resist the crash and
burn instinct?
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
"Language is a virus from outer space" - William S. Burroughs
Steve Tynor
Georgia Tech Research Institute
tynor@prism.gatech.edunazgul@alphalpha.com (Kee Hinckley) (12/12/90)
In article <18470@hydra.gatech.EDU> tynor@prism.gatech.EDU (Steve Tynor) writes: >I always happens shortly after I start a 'make' in an xterm. I suspect that the >machine is running out of process table, but that's just a guess. This is Why is it a guess? What does a "ps" show? If you are getting anywhere past 40 then watch out, you're too close to the top. I used to crash regularly due to running sendmail and a mailing list (which reinvoked sendmail from a perl script). I finally switched to batching mail, which cut down on my problems considerably. Running out of disk space is another option, X and Emacs can take up a fair amount. -- Alphalpha Software, Inc. | motif-request@alphalpha.com nazgul@alphalpha.com |----------------------------------- 617/646-7703 (voice/fax) | Proline BBS: 617/641-3722 I'm not sure which upsets me more; that people are so unwilling to accept responsibility for their own actions, or that they are so eager to regulate everyone else's.
tynor@prism.gatech.EDU (Steve Tynor) (12/13/90)
In article <1990Dec12.060416.12660@alphalpha.com> nazgul@alphalpha.com (Kee Hinckley) writes: >In article <18470@hydra.gatech.EDU> tynor@prism.gatech.EDU (Steve Tynor) writes: >>I always happens shortly after I start a 'make' in an xterm. I suspect that the >>machine is running out of process table, but that's just a guess. This is > >Why is it a guess? What does a "ps" show? If you are getting anywhere >past 40 then watch out, you're too close to the top. I used to crash Because make was failing with a 'unable to fork process' message and the crash code was: 00040010: "invalid segment count (OS/MST manager)" (not that I have any idea what an invalid segment count is). I just did a 'ps aux|wc' and got 39 processes. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= "Language is a virus from outer space" - William S. Burroughs Steve Tynor Georgia Tech Research Institute Artificial Intelligence Branch tynor@prism.gatech.edu
tomg@hpcvlx.cv.hp.com (Thomas J. Gilg) (12/14/90)
> Because make was failing with a 'unable to fork process' message and the crash > code was: 00040010: "invalid segment count (OS/MST manager)" (not that I have > any idea what an invalid segment count is). Sounds like you're hitting the process limit. I do it all the time :-( Thomas Gilg tomg@cv.hp.com