wdqs-updater stopped processing updates on wdqs1002 on August 25, from ~10:20 to 10:40 (UTC). Looking at the logs, there is no obvious errors except the RecentChangePoller "getting too close to the current stream" and a failed HTTP request just before the updater starts processing updates again. See P5920 for an extract of the logs.
Description
Description
Related Objects
Related Objects
- Mentioned Here
- P5920 wdqs1002 failed updater
Event Timeline
Comment Actions
Similar issue happened again, between 5:50am and 8:50am UTC today (Sept 1). Again, updater recover after an HTTP exception. See logstash for details.
Comment Actions
This seems to be the trigger:
Aug 25 10:35:25 wdqs1002 bash[18627]: 10:35:25.238 [main] INFO o.w.query.rdf.tool.rdf.RdfRepository - HTTP request failed: java.util.concurrent.ExecutionException: java.lang.IllegalArgumentException: Buffering capacity exceeded, attempt 1, will retry
but I have no idea what that message means... This SO article provides some clues, but I'm not sure why RdfRepository produces overly large answer (over 2M) - generally we don't need something from it that is so large.
Comment Actions
I can't find any other instance of such exception in the logs, so I'm closing this for now, since I can't really find the cause without recent occurrence in the logs.