home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!uwm.edu!caen!batcomputer!lynx.msc.cornell.edu!ransom.msc.cornell.edu!rick
- From: rick@ransom.msc.cornell.edu (Rick Cochran)
- Newsgroups: comp.unix.aix
- Subject: Re: runaway processes (was Re: Vi is still broken)
- Message-ID: <1992Aug19.132117.5939@msc.cornell.edu>
- Date: 19 Aug 92 13:21:17 GMT
- Article-I.D.: msc.1992Aug19.132117.5939
- References: <1992Aug17.163739.29534@APS.Atex.Kodak.COM> <133789@lll-winken.LLNL.GOV>
- Sender: news@msc.cornell.edu
- Organization: Cornell Materials Science Center
- Lines: 34
-
- In article <133789@lll-winken.LLNL.GOV>, terry@phoenix.ocf.llnl.gov (Terry Heidelberg x24154) writes:
- |> We don't see hangs; what we see is an occasional interactive program
- |> looping to beat the band. Since upgrading to 3.2.1 (later to 3.2.2)
- |> we have seen the following programs looping:
- |> vi aixterm less info_ascii
- |> and maybe a few more I forget. The typical scenario is the user
- |> telnet's in, works away, and for some reason has the session broken
- |> (who knows, maybe after something hangs) after which s/he logs in
- |> again, or goes to some other work, unaware that one of his/her processes
- |> is running amok. Since we have 18 550 cycle servers to
- |> look after, some of these runaways chew up quite a bit of cpu time
- |> before they get killed by a system administrator. I have vague
- |> suspicions of (a) the telnetd, (b) something in the way these programs
- |> handle signals or system-call errors, (c) [the ever popular] AIX.
- |>
- |> Anyone have similar experiences, or ideas about the cause? Are the
- |> above fixes likely to help us?
- |>
- |> Thanks.
-
- Yup. We see this kind of thing frequently. Since we are about to start
- chargeback accounting, this will be a severe pain.
-
- Our most recent method of producing the problem is to close a window in
- which a process which is on the receiving end of a pipe is running.
-
- This has been going on for years but has lately become intolerable. I
- must now try to come up with a reproducible example and phone up IBM
- to see if it's a 'supported defect' :-}
-
- --
- |Rick Cochran 607-255-7223 rick@msc.cornell.edu|
- |Cornell Materials Science Center rick@crnlmsc2.bitnet|
- |E20 Clark Hall, Ithaca, N.Y. 14853 cornell!msc.cornell.edu!rick|
-