RFR : JDK-8141591 - javax/management/remote/mandatory/threads/ExecutorTest.java fails intermittently

Daniel Fuchs daniel.fuchs at oracle.com
Mon Nov 14 10:44:27 UTC 2016


On 14/11/16 06:57, Harsha Wardhana B wrote:
>>> Well, that will not cover the case where user shuts down executor but
>>> keeps the client/server alive. So we will need a executor that can keep
>>> notif system running as well as do clean-up if client/server is closed.
>>
>> It's OK to continue in order to do clean up and
>> shutdown gracefully. It seems wrong to keep going afterwards
>> as if nothing had happened though (in the case the
>> user shuts the supplied executor down).
> With current changes, we do continue to carry on with linear executor.
> If the user wants to shutdown the system, he can always do it by
> shutting down client and server along with executor. If he shuts down
> executor but not client/server, it may be safe to assume that he wants
> the system to be up and running. It may not be appropriate to assume
> user wants to shutdown notif system just because he shutdown executor.

Well, it may also not be appropriate to invoke a user provided callback
on another executor than the one provided by the user.
If the user provides an executor, we must assume he has good
reasons to do so.
I'm not hard set to opposing to what you propose, but it makes me
feel uncomfortable.

best regards,

-- daniel


More information about the serviceability-dev mailing list