Bug 737044 - Investigate deferred link processing in the case of "agent unknown"
Summary: Investigate deferred link processing in the case of "agent unknown"
Alias: None
Product: Red Hat Enterprise MRG
Classification: Red Hat
Component: cumin
Version: Development
Hardware: Unspecified
OS: Unspecified
Target Milestone: ---
: ---
Assignee: grid-maint-list
QA Contact: MRG Quality Engineering
Depends On:
TreeView+ depends on / blocked
Reported: 2011-09-09 13:02 UTC by Trevor McKay
Modified: 2016-05-26 20:14 UTC (History)
2 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Last Closed: 2016-05-26 20:14:37 UTC

Attachments (Terms of Use)

Description Trevor McKay 2011-09-09 13:02:43 UTC
Description of problem:

This is a tracker BZ.  On a very slow network (from home) a case was seen in which a jobserver object referenced a scheduler object, but the agent associated with the scheduler had been deleted by the time the jobserver update was processed (because of a timeout?).  Deferred link processing for the reference to the scheduler object was skipped, because the check for the agent failed.  This happened cyclically.

As a consequence, no submissions were ever displayed.  Even if the scheduler object was cyclically deleted, having submissions come and go on the screen would at least be a visual cue of a slow network (better than just blank forever).

Since the objId field contains the name of the agent in the reference, perhaps we can extend deferred link processing to include the unknown agent case, with some expiration date on deferred links spawned by unknown agents.  If we did this, when the scheduler object eventually was recreated submissions would have appeared (until the next timeout and delete).

Version-Release number of selected component (if applicable):

How reproducible:

Steps to Reproduce:
Actual results:

Expected results:

Additional info:

Comment 2 Anne-Louise Tangring 2016-05-26 20:14:37 UTC
MRG-Grid is in maintenance and only customer escalations will be considered. This issue can be reopened if a customer escalation associated with it occurs.

Note You need to log in before you can comment on or make changes to this bug.