Closed pavelhoral closed 1 year ago
This might be issue only for embedded DS (i.e. not a production issue). Seems it is caused by asynchronous processing of the shutdown hook - admin token is being destroyed after CTS store (embedded DS) connection processing is closed.
Just tested with external DS instance and the race condition with CTS is still present:
Sometimes the shutdown completes and the following exception is logged to standard output:
Some additional information:
There are two questions regarding what is happening:
And also the correct solution might be to actually shutdown executor services as the last step... btw. ExtendedExecutorServiceFactory
is marked as temporary in its JavaDoc and that might be connected to the issue as well.
The issue has been fixed through #95.
When I shut down docker image (#61) it hangs indefinitely due to the following thread:
I have not invested any time in investigating the root cause yet.