Hide Forgot
Description of problem: Using Red Hat High Availability to manage an ha schedd on a CM produces a problem because other processes, like the negotiator, need to access SPOOL. Add params SCHEDD.SPOOL and SCHEDD.HISTORY. Create SharedJobQueueLocation feature: - SCHEDD.SPOOL: 0 - SCHEDD.HISTORY: $(SCHEDD.SPOOL)/history Modify BaseHAScheduler: - remove JobQueueLocation from include - add SharedJobQueueLocation to include Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Technical note added. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. New Contents: C: Many components of MRG Grid rely on SPOOL C: The schedd can't use a SPOOL defined specifically for the schedd C: Added SCHEDD.SPOOL and SCHEDD.HISTORY R: The schedd can use a SPOOL defined specifically for the schedd
Tested on RHEL6.2 x x86_64/i386 with condor-classads-7.6.5-0.8.el6 condor-plumage-7.6.5-0.8.el6 condor-ec2-enhanced-hooks-1.2-4.el6 python-condorutils-1.5-4.el6 condor-job-hooks-1.5-4.el6 condor-ec2-enhanced-1.2-3.el6 condor-wallaby-tools-4.1.2-1.el6 condor-wallaby-base-db-1.17-2.el6 condor-debuginfo-7.6.5-0.8.el6 condor-7.6.5-0.8.el6 condor-kbdd-7.6.5-0.8.el6 condor-aviary-7.6.5-0.8.el6 condor-qmf-7.6.5-0.8.el6 python-condorec2e-1.2-4.el6 condor-wallaby-client-4.1.2-1.el6 ricci-0.16.2-43.el6 cman-3.0.12.1-23.el6 rgmanager-3.0.12.1-5.el6 and there is no error in logfiles and condor_history/_q/_status works also with different values of SPOOL and *.SPOOL.