[Imported from gridengine issuezilla ​http://gridengine.sunsource.net/issues/show_bug.cgi?id=3186]

        Issue #:      3186             Platform:     All      Reporter: petrik (petrik)
       Component:     gridengine          OS:        All
     Subcomponent:    install          Version:      6.2u4       CC:    None defined
        Status:       NEW              Priority:     P1
      Resolution:                     Issue type:    DEFECT
                                   Target milestone: ---
      Assigned to:    dom (dom)
      QA Contact:     dom
          URL:
       * Summary:     Updating to 6.2u4 causes old running and pending jobs to be "lost"
   Status whiteboard:
      Attachments:

     Issue 3186 blocks:
   Votes for issue 3186:


   Opened: Thu Nov 19 05:00:00 -0700 2009 
------------------------


When there are running jobs before update to 6.2u4 is started. These jobs will not be seen by qmaster and will be missing in the
accounting/reporting file after they finish. This is a regression.

Both qmaster/execd messages file will contain for each such job line on new qmaster/execd startup similar to:
11/19/2009 10:59:20|  main|<host>|E|format error while unpacking job "jobs/00/0000/0003"

   ------- Additional comments from reuti Thu Nov 19 05:12:41 -0700 2009 -------
Confirmed - http://gridengine.sunsource.net/ds/viewMessage.do?dsMessageId=224667&dsForumId=38

I wasn't sure, whether the update step was too big.

   ------- Additional comments from petrik Mon Nov 23 06:27:27 -0700 2009 -------
Also old pending jobs are "lost" in any setup.

[Imported from gridengine issuezilla ​http://gridengine.sunsource.net/issues/show_bug.cgi?id=249]

        Issue #:      249             Platform:     All           Reporter: petrik (petrik)
       Component:     testsuite          OS:        All
     Subcomponent:    tests           Version:      current          CC:    None defined
        Status:       RESOLVED        Priority:     P3
      Resolution:     DUPLICATE      Issue type:    ENHANCEMENT
                                  Target milestone: milestone 1
      Assigned to:    issues@testsuite
      QA Contact:     joga
          URL:
       * Summary:     Need test for issue 2694
   Status whiteboard:
      Attachments:

     Issue 249 blocks:
   Votes for issue 249:     Vote for this issue


   Opened: Mon Oct 20 07:33:00 -0700 2008 
------------------------


Need test for issue 2694

   ------- Additional comments from petrik Mon Oct 20 07:34:34 -0700 2008 -------
duplicate

*** This issue has been marked as a duplicate of 237 ***

I have installed SGE 8.1.9 from RPM on SUSE 11 SP4.

PROBLEM: Running this command hangs with no output.

sudo /sbin/service sgeexecd.p6444 status

DIAGNOSIS: The cause of this is the sgeexecd script failing to set pidfile and then calling cat which now expects STDIN and so waits indefinitely.

FIX: This is the patch that I used on the script in /etc/init.d. It is a two line change that you can apply to the appropriate source file.

--- sgeexecd.p6444.orig 2016-11-18 17:42:28.000000000 +0000 +++ sgeexecd.p6444 2016-11-22 15:16:04.000000000 +0000 @@ -448,9 +448,10 @@

fi

if [ "$status" = true ]; then

+ pidfile=$execd_run_dir/execd.pid

if [ -f $pidfile ]; then

pid=cat $pidfile

• if $utilbin_dir/checkprog $pid $name > /dev/null; then

+ if $utilbin_dir/checkprog $pid sge_execd > /dev/null; then

echo "execd (pid $pid) is running..." exit 0

else

VERIFY: Applying that patch causes the status subcommand to report correctly.

Regards,

Pete Forman Senior Programmer

Dolphin Geophysical Limited A member of the Shearwater GeoServices? Group

Direct Dial: +44 (0) 1892 707173 / Mobile: +44 (0) 7840 797658 / pete.forman@… Brockbourne House, 77 Mount Ephraim, Tunbridge Wells, Kent, TN4 8BS, UK www.shearwatergeo.com