Details
-
Bug
-
Resolution: Done
-
Major
-
None
-
None
Description
There is a situation, where it could happen that a latch is not being released forever. It could happen on the second request to a page, if the first request has not been finished and put to cache while the second goes through the CacheReadValve. But if the first request finishes right after that - and before the second request is checking the generatingPage-hashmap in the EngineValve - the second request will open a new latch (as the first request has finished already), but right after that it will read the result of the first entry from cache without afterwards releasing the latch. As long as the affected page is served from cache, the problem is not seen, only when it has to be rendered again for the same user, workflow, language,...