Closed pabender closed 6 years ago
This is mostly solved by #122.
Our tests now clean most of their own threads. JMRI tests produce a significantly fewer number of threads, but the number is still several hundred.
This has been reduced sufficiently that this issue will be closed. Most of the still open threads are waiting to be garbage collected when tests complete.
Thank you for your work on this issue, and also to others for their input.
David
On Sun, Apr 29, 2018, 06:07 Paul Bender notifications@github.com wrote:
Closed #119 https://github.com/openlcb/OpenLCB_Java/issues/119.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/openlcb/OpenLCB_Java/issues/119#event-1600518205, or mute the thread https://github.com/notifications/unsubscribe-auth/AAg4SsRGijA_tLBZBmcp8O9e8Dbf66dHks5ttbsOgaJpZM4RvjkU .
This is primarily a testing issue, but when OpenLCB objects are created, there is no way to cleanly destroy the threads they leave behind. We need to find a way to do this.
This was discovered while working on a JMRI issue ( https://github.com/JMRI/JMRI/pull/4818 ) where the JMRI test suite creates a very large number of OpenLCB related objects in testing its interfaces, but has no way to terminate the threads left behind.