Opened 50 years ago

Last modified 9 years ago

#894 new defect

IZ588: gridengine adapter does not support permanent request slo

Reported by: easymf Owned by:
Priority: low Milestone:
Component: hedeby Version: 1.0u2
Severity: Keywords: gridengine_adapter
Cc:

Description

[Imported from gridengine issuezilla http://gridengine.sunsource.net/issues/show_bug.cgi?id=588]

        Issue #:      588                      Platform:     All         Reporter: easymf (easymf)
       Component:     hedeby                      OS:        All
     Subcomponent:    gridengine_adapter       Version:      1.0u2          CC:    None defined
        Status:       NEW                      Priority:     P4
      Resolution:                             Issue type:    DEFECT
                                           Target milestone: 1.0u5next
      Assigned to:    rhierlmeier (rhierlmeier)
      QA Contact:     rhierlmeier
          URL:
       * Summary:     gridengine adapter does not support permanent request slo
   Status whiteboard:
      Attachments:


     Issue 588 blocks:
   Votes for issue 588:     Vote for this issue


   Opened: Fri Oct 31 03:36:00 -0700 2008 
------------------------


   Description:

   When an user configures GE adapter to use PermanentRequestSLO, GE adapter does
   not know how to handle it - there is a message in the log file:

   10/30/2008 11:13:51|13|E|Do not know howto handle SLO PermanentRequestSLOge
   (type PermanentRequestSLOConfig)

   A user can hit this problem when a GE adapter configuration like this is used:

   <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
   <common:componentConfig xsi:type="ge_adapter:GEServiceConfig"
                           mapping="default"
                           xmlns:executor="http://hedeby.sunsource.net/hedeby-execu
   tor"
                           xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
                           xmlns:reporter="http://hedeby.sunsource.net/hedeby-repor
   ter"
                           xmlns:security="http://hedeby.sunsource.net/hedeby-secur
   ity"
                           xmlns:resource_provider="http://hedeby.sunsource.net/hed
   eby-resource-provider"
                           xmlns:common="http://hedeby.sunsource.net/hedeby-common"

                           xmlns:ge_adapter="http://hedeby.sunsource.net/hedeby-gri
   dengine-adapter">
       <common:slos>
           <common:slo xsi:type="common:PermanentRequestSLOConfig"
                       urgency="50"
                       name="permanent_request"/>
       </common:slos>
       <ge_adapter:connection keystore="/var/sgeCA/port1978/default/../keystore"
                              password="changeit"
                              username="username"
                              jmxPort="9854"
                              execdPort="1979"
                              masterPort="1978"
                              cell="default"
                              root="/cod_home/.../ts_dist/ge_0"
                              clusterName="p1978"/>
       <ge_adapter:sloUpdateInterval unit="minutes"
                                     value="1"/>
       <ge_adapter:jobSuspendPolicy suspendMethods="reschedule_jobs_in_rerun_queue
   reschedule_restartable_jobs suspend_jobs_with_checkpoint">
           <ge_adapter:timeout unit="minutes"
                               value="2"/>
       </ge_adapter:jobSuspendPolicy>
       <ge_adapter:execd adminUsername="root"
                         defaultDomain=""
                         ignoreFQDN="true"
                         rcScript="false"
                         adminHost="true"
                         submitHost="true"
                         cleanupDefault="false">

   <ge_adapter:localSpoolDir>/usr/local/testsuite/1978/execd</ge_adapter:localSpoolDir>
       </ge_adapter:execd>
   </common:componentConfig>

   Evaluation:

   A lower priority issue as the system remains functional even if wrong SLO is
   used. There is also a quite easy workaround.

   Suggested fix/ Work Around:

   Workaround - instead of PermanentRequestSLO (that is never met) it is possible
   to use MinResourceSLO that asks for reasonably big amount of resources.

   Fix - add support of PermanentRequestSLO to GE adapter


   Analysis:

   The problem is caused by fact, that PermanentRequestSLO is not supported by
   default SLO factory nor by GE adapter SLO factory.

   The corresponding classes are com.sun.grid.grm.service.slo.SLOFactory and
   com.sun.grid.grm.service.impl.ge.GEServiceSLOFactory.

   How to test:

   Both junit and TS test have to created:
   1. junit test has to test if a GEServiceSLOFactory is able to create a
   PermanentRequestSLO from a given PermanentRequestSLOConfig instance
   2. TS test has to test that GE adapter is able to be started with configuration
   containing PermanentRequestSLO, that it is functional and that no error message
   (like described by the issue) is in the log file.

   ETC:

   0,5 PD fix
   0,5 PD junit test
   2,0 PD TS test

   ATC:
               ------- Additional comments from rhierlmeier Fri Dec 5 02:39:34 -0700 2008 -------
   The issue has been fixed in the Hedeby code with issue 605. A junit test is also
   available with check in of issue 605.

   Only the testsuite test is missing.
               ------- Additional comments from rhierlmeier Wed Nov 25 07:21:10 -0700 2009 -------
   Milestone changed

Change History (0)

Note: See TracTickets for help on using tickets.