SUMMARY: 5.1a pk#5 and brackets around 'ps' process output

From: Mike Broderick (broderic_at_MIT.EDU)
Date: 09/08/03

  • Next message: McCracken, Denise: "Can't find ."
    Date: Mon, 08 Sep 2003 11:58:03 -0400
    To: tru64-unix-managers <tru64-unix-managers@ornl.gov>
    
    

    I had many replies to this, some some echoing the standard "don't reply
    on ps output, use a PID file" that was good advice but not applicable
    lest I wanted to hack commercial third party scripts. Some recommended
    different ps command switches but nothing I tried changed the output in
    the CMD field (still truncated and in brackets).

    The solution that worked that a few recommended was calling HP s/w
    support to obtain/install a customer specific patch
    (T64KIT0019721-V51AB23-20030822 for 5.1a pk#5).

    Thanks to Jenny Butler, Debra Alpert, Steve Herber, Phillip Brown,
    Jean-marc Vincent, Bob Vickers, and Chris Medaglia for their replies

                                                                                                           
    _Mike

    Mike Broderick wrote:

    > We just applied patch kit #5 to a 5.1a system to fix a problem and now
    > several processes (all the Netbackup ones) show up in a 'ps' listing
    > with brackets around the command (and command paths and params no
    > longer appear). E.g.,
    >
    > pk#5:
    >
    > root 877 1 0.0 14:18:19 ?? 0:00.55 [ltid]
    > root 883 1 0.0 14:18:19 ?? 0:00.31 [bprd]
    > root 885 1 0.2 14:18:19 ?? 0:06.03 [vmd]
    > root 889 1 0.0 14:18:19 ?? 0:01.04 [bpdbm]
    > root 985 877 0.0 14:18:23 ?? 0:00.03 [acsd]
    > root 1038 877 0.0 14:18:25 ?? 0:00.03 [tldd]
    > root 1095 877 0.0 14:18:27 ?? 0:08.37 [avrd]
    > root 1096 985 0.0 14:18:27 ?? 0:00.01
    > [acssel]
    >
    > pk#4:
    >
    > root 439603 440795 0.0 Aug 26 ?? 2:39.99 avrd
    > root 440722 1 0.0 Aug 26 ?? 2:26.44 vmd -v
    > root 440765 1 0.0 Aug 26 ?? 0:00.00 portmap
    > root 440767 1 0.0 Aug 26 ?? 0:09.50
    > /usr/openv/netbackup/bin/bprd
    > root 440795 1 0.0 Aug 26 ?? 0:12.70
    > /usr/openv/volmgr/bin/ltid
    > root 440820 1 0.0 Aug 26 ?? 0:01.49 tldcd -v
    > root 440833 440891 0.0 Aug 26 ?? 0:00.17 acsssi 13741
    > root 440842 440795 0.0 Aug 26 ?? 0:00.58 tldd -v
    > root 440853 440891 0.0 Aug 26 ?? 0:00.01 acssel -s 13740
    >
    >
    > /usr/bin/ps was patched but the old 'ps' also shows the changed output.
    >
    > This is breaking several Netbackup scripts that apparently rely on
    > that output. I browsed the 119 patch list but could not readily
    > identify the one that made this change.
    >
    > Anyone else seen this or know how to undo just this change.
    >
    >
    > _Mike
    >


  • Next message: McCracken, Denise: "Can't find ."

    Relevant Pages

    • Re: YAFM: The lich casts a spell at you! Your pair of leather gloves corrodes!
      ... (so, e.g., foosh and L's replies to this thread are missing ). ... of the page definitely omits anything in brackets. ... newsgroup header lines at the top. ... subjects in threading order) but the titlebar doesn't show the ...
      (rec.games.roguelike.nethack)
    • SUMMARY: Show all NFS Mounts
      ... I even got the Out of Office replies!! ... I am taking down server1 which shares out directory /a and directory /b. ... mounted directories FROM server1. ... I know I have done this before, but I cannot remember the command. ...
      (SunManagers)
    • Re: Mail / Conversation Threading in Alpine / Pine?
      ... I find that all of the messages in a conversation, including my replies, within a folder are properly threaded. ... I use the S command to move all of the messages into the same folder so that they can be threaded. ... Or it could happen if the user creates a new message instead of using the reply function, ie. the user is incompetent. ...
      (comp.mail.pine)
    • file server slow to respond
      ... (please cc: replies) ... Logging in (ssh) takes several seconds. ... of the command occur quickly and normally. ...
      (Debian-User)
    • Re: MuPad 3.0 extraction of individual eigenvectors
      ... I have read your replies with a great deal of interest. ... It is often the case that a user of MuPad does not want to ... simply take output from a command to MuPad, but the user often wants do ... the required techniques are sprinkled in amongst the existing documentation, ...
      (sci.math.symbolic)