Hi. You can find stacktrace from 2 nodes while destroying cache. Our primary worker threads are AsyncWorker-*. For about 40 seconds between 24/Oct/2018 15:52:51 - 24/Oct/2018 15:53:28 threads waiting for cache to be destroyed.
As a note I can say that is not an issue every time. On Thu, Oct 25, 2018 at 6:09 PM akurbanov <[email protected]> wrote: > Hello, > > I would definitely suggest you to upgrade to latest Ignite versions as a > vast number of issues has been fixed since 1.9: > https://issues.apache.org/jira/browse/IGNITE-6035 > https://issues.apache.org/jira/browse/IGNITE-2714 > https://issues.apache.org/jira/browse/IGNITE-8531 > https://issues.apache.org/jira/browse/IGNITE-8021 > https://issues.apache.org/jira/browse/IGNITE-7319 > > Please pay attention to the last one, this seems to be exactly your use > case. A lot of things happen on both node and cluster levels when you call > cache.destroy(), since this is a minor exchange which involves all server > nodes where this cache exists, also a checkpoint along with write-ahead log > flush is performed. > > Anyway, could you please give more details on how exactly your issue is > meant to be reproduced (check the code on IGNITE-7319 if it is looks like > your case), provide some logs and thread dumps to analyze. > > Regards, > Anton > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ > -- Alper Tekinalp Software Development Team Lead Evam Streaming Analytics www.evam.com <http://www.evam.com.tr> <http://www.evam.com>
