MODE-1465 - Added additional event for sequencing failures #364
Conversation
| stats.recordDuration(DurationMetric.SEQUENCER_EXECUTION_TIME, durationInNanos, TimeUnit.NANOSECONDS, payload); | ||
| } | ||
| } catch (Throwable t) { | ||
| fireSequencingFailureEvent(selectedNode, inputSession, t); |
There was a problem hiding this comment.
Do we capture the sequencer name, sequenced path, and the output path in the failure event? This is all useful information, especially when a sequencer generally works but fails on particular inputs.
There was a problem hiding this comment.
sequencer name - no, sequenced path - yes,
the output path not sure - should it be the configured output path ?
There was a problem hiding this comment.
Really, we should probably have all the same information in the failure event as in the success event, except the failure should also include a useful representation of the cause of the failure. You'll notice the old log statements did output information from the WorkItem; I think we should include that information in the event.
There was a problem hiding this comment.
in the successful case, we have the list of output nodes which were created. In the case of an unexpected exception, how can we get that (the exception can occur at any point during the sequencing code and can have any number of causes) ?
There was a problem hiding this comment.
True, the output nodes may not be available, but we certainly know the output path(s) we passed to the sequencer. Isn't that useful information?
There was a problem hiding this comment.
yup, sure it is. Atm the jcr event has the Throwable cause, but i'll update the pull request and add the extra information from the work item, as you suggested.
…quencer name, output path and selected path (from the sequencer work info)
|
Updated the information supplied to sequencing events with user id, sequencer name, output path and selected path (from the sequencer worker) |
|
Looks great. Thanks, @hchiorean! Will merge shortly. |
|
Merged into the 'master' branch after fixing a (minor) merge conflict. |
No description provided.