home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.sun.admin
- Path: sparky!uunet!news.larc.nasa.gov!cscsun2.larc.nasa.gov!dan
- From: dan@cscsun2.larc.nasa.gov (Dan O'Brien)
- Subject: Runaway portmap on a Sun3/280 SunOS 4.0.3 (Proc: table full)....
- Message-ID: <Bs2A0C.EBo@news.larc.nasa.gov>
- Keywords: portmap NFS Sun3/280
- Sender: news@news.larc.nasa.gov (USENET Network News)
- Organization: NASA Langley Research Center, Hampton, VA USA
- Date: Mon, 27 Jul 1992 18:57:48 GMT
- Lines: 25
-
- We have a Sun3/280 running SunOS 4.0.3. For a while now I've
- been getting
- "proc: table full"
- messages followed by a system lockup (can't log in and can't start
- new processes if you are logged in). This is obviously a situation
- where we have too many processes and I bumped up the max processes
- and installed a new kernel. The problem continued and I eventually
- started analyzing the kernel core dump. I'm fairly new at using
- adb so I was only able to ascertain that the runaway process was
- "portmap". It appears to just spawn copies of itself
- off until the process table is full. I tried applying all the
- relevant NFS patches I could find for 4.0.3 and the problem
- disappeard...for a while. Now it's back again and I don't know
- where to turn. Have I missed a patch somewhere? Is my NFS
- exports to other Sun3's, an SGI, and a SparcStation 2 causing
- me problems? I suspect the Sparc2 as a possible culprit
- (It's running 4.1.1). Upgrading to 4.1.1 on the Sun3/280 is not
- an option at this point. Any help such as pointers to appropriate
- patches would be greatly appreciated.
-
-
- Dan O'Brien
- NASA Langley Research Center
- dan@cscsun2.larc.nasa.gov
- dan%cscsun2.larc.nasa.gov@uunet.uu.net
-