No there is nothing apart from the log in the event admin implementation. But if you're dealing with Sling jobs you shouldn't use the event admin way of starting/processing jobs anyway. It's prefered to use the api instead (JobManager and JobConsumer)
Carsten 2014-08-22 15:19 GMT+02:00 Felix Meschberger <[email protected]>: > Hi Tim > > Blacklisting is a feature of the Apache Felix Event Admin bundle and IIRC > at least the timeout for blacklisting is configurable. You might want to > check your application, whether blacklisting timeout has been reconfigured > or not. > > I am not really aware, whether it can be analyzed whether handlers have > been blacklisted or not. Other than inspecting the log file for respective > entries. > > Regards > Felix > > Am 22.08.2014 um 10:28 schrieb Timothée Maret <[email protected]>: > > > Hi, > > > > In our deployment we have experienced tens of thousands of Sling Job > > pilling up in queue without having them being processed (for days). > > Using the regular Sling Jobs UI, we found no reason why this happened > > (queue was configured properly and active). > > We also analyzed multiple thread dumps and found no deadlock that could > > explain this issue. > > > > Quoting the doc [0] it may have been that our job topics where simply > > blacklisted: > > > > "The OSGi EventAdmin usually comes with a timeout for event handlers. An > > event handler must consume an OSGi event as fast as possible otherwise > the > > handler might get a timeout and get blacklisted" > > > > How can we check for blacklisted handlers ? > > In which bundle is the blacklisting feature implemented ? > > > > Regards, > > > > Timothee > > > > [0] > > > https://sling.apache.org/documentation/bundles/apache-sling-eventing-and-job-handling.html > > -- Carsten Ziegeler Adobe Research Switzerland [email protected]
