Very Computer

sh:  fork failed, too many processes

I checked the kernel setting for the maximum number of processes and
bumped it up to 200, relinked the kernel, took the system down and
back up, and the message reappeared.

I then ran sa1 then sar -v, and it showed that there were 78 processes
running out of a maximum allowable 200.  I did notice an unusual
number of mail processes owned by ? when I ran ps -ef, but I looked
in the cron tables and found nothing unusual.

Has anyone seen this error before?  If so, how did you fix it?

ps.  I noticed SCO took the knowledgebase text search feature of
sos.sco.com offline, where you used to be able to search for things
like this.

 I guess the greedy bastards (excuse my choice of words but that's my
impression of them after dealing with them a while) felt that it
wasn't cost effective for them to allow users to do something that
other, bigger companies like Microsoft offer for free, any longer.  I
also noticed that they are now offering a subscription for  a fee for
this service on their Website.  I hope this strategy backfires on
them.  Even though I love Unix, I can see why NT is gaining market
share, with the arrogant attitude SCO shows towards its customers.

Quote:jeffrey stewart writes:
> ps.  I noticed SCO took the knowledgebase text search feature of
> sos.sco.com offline, where you used to be able to search for things
> like this.

>  I guess the greedy bastards (excuse my choice of words but that's my
> impression of them after dealing with them a while) felt that it
> wasn't cost effective for them to allow users to do something that
> other, bigger companies like Microsoft offer for free, any longer.  I
> also noticed that they are now offering a subscription for  a fee for
> this service on their Website.  I hope this strategy backfires on
> them.  Even though I love Unix, I can see why NT is gaining market
> share, with the arrogant attitude SCO shows towards its customers.

Right now, in two X windows, I'm looking at TWO search engines for
SCO knowledgebase articles, each of which I reached with a few clicks
off the www.sco.com page:

        http://sco.netpub.com/cgi-bin/webinator
        http://www.sco.com/Support/ssl.html

--

: ps.  I noticed SCO took the knowledgebase text search feature of
: sos.sco.com offline, where you used to be able to search for things
: like this.

Try:  http://www.sco.com/Support/ssl.html
for the allegedly missing text search feature.

:  I guess the greedy bastards (excuse my choice of words but that's my
(... ad nausium)
: I also noticed that they are now offering a subscription for a fee for
: this service on their Website.  I hope this strategy backfires on
: them.  Even though I love Unix, I can see why NT is gaining market
: share, with the arrogant attitude SCO shows towards its customers.

Really?  Where did you see that?
Just a minor reminder:  At one time, SCO charged for access to
their bbs.  This was before the world discovered the internet.
The price was about $99/year which I gladly paid.  It's now free
along with web, ftp and uucp access.

Public appologies graciously accepted.

[x] Email to author  [ ] To mailing list  [x] Posted to newsgroup
--
# Jeff Liebermann  Liebermann Design  150 Felker St #D  Santa Cruz  CA  95060

# 408.699.0483 digital_pager    73557,2074  cis [don't]

Basically SCO keeps a process open for a hidden or dead
process.  Worse, it often keeps more than one process
open.

You can go in using SCOSH all kill all unatached
processes, but this is boring as sin.

Its easier just to shut the server down once a week.

Its not much, but it works for us.

scn

  Just a hunch, but it sounds like mmdf may be mailing itself
to death.  If you look in the /usr/spool/mmdf/lock/home/*
directories, you may notice you cannot get a file listing
due to an incredible number of entries.  Take a look at your
system, and E-mail me if you find this is the case.  I can
give you some pointers on what to do about this, if you need
them.  

   Matthew Byrd