FELIX-6784 Reduce number of thread (re)creations of the Configurator Worker Thread #427
Conversation
| } | ||
| } | ||
| }); | ||
| setKeepAliveTime(5, TimeUnit.SECONDS); |
There was a problem hiding this comment.
This could be made configurable, but my initial proposal is to have a static, relatively short, keep alive time fixed in the code. Its purpose is mainly to let the underlying thread terminate at some point if no tasks are queued.
|
Could you have a look at the requested changes @tjwatson ? |
|
Hi @stbischof & @tjwatson , could you have a look at this PR? |
|
Hi, did you run this against the tck? Can you rebase your pr and push again? Then some actions should start the test. |
… ScheduledThreadPoolExecutor. This already implements keeping track of tasks, and allows for the configuring a keep alive time.
glimmerveen
force-pushed
the
feature/FELIX-6784
branch
from
4845775 to
e184e7c
Compare
I did not run the TCK myself, as I don't know how to do this; is there instructions on how to do this? I did rebase the changes and push the change. |
|
i see, so we have to add tck for configurator first. do you know how to? |
I do see a way to add configurator's regular maven build to the ci workflows, but I don't really grasp how I could include the TCK for configurator specifically. I see there is a framework.tck module, but it does not build on my machine so I am not really sure how I could use that to setup an equivalent setup for configurator. |
2 participants
Changed the WorkerQueue to be based on the Java's standard ScheduledThreadPoolExecutor. This already implements keeping track of tasks, and allows for the configuring a keep alive time.