Bugzilla – Bug 2578
reliable state change notification
Last modified: 2012-09-05 11:42:16
You need to log in before you can comment on or make changes to this bug.
For reliable state change notifications from each jobs submitted to the MJFS/MEJS, the MMJS should implement fall-back RP polling for all state monitoring. this is to ensure that a state change is detected even in the case when a notification was not recieved by the MMJS.
Looks like I am obtaining the same kind of weird result. The Java rendezvous unit tests did not feature a testing of GetResourceProperty or GetMultipleResourceProperties and thus did not put this issue into evidence (the registrant data as embedded in the notifications message still seems correct according to a re-run of the existing Java tests). ). I am looking more into this. I have had issues with Core in the past setting RPs that are Java arrays - maybe my work around of the time is now generating a bug due to changes in Core? I will try a few things.
PLEASE IGNORE PREVIOUS POST
Frankly I think the title of this bug suggests what a real fix would be for this issue. We should simply have a reliable version of notifications--a two way message that gives an ACK back to the notification producer. Implementing polling is simply saying "we're working around a feature that hasn't been developed but that we need".
Reassigning to current GRAM developer to close/fix as appropriate.
Doing some bugzilla cleanup... Resolving old GRAM3 and GRAM4 issues that are no longer relevant since we've moved on to GRAM5. Also, we're now tracking issue in jira. Any new issues should be added here: http://jira.globus.org/secure/VersionBoard.jspa?selectedProjectId=10363
