zentol commented on pull request #15054: URL: https://github.com/apache/flink/pull/15054#issuecomment-810153918
> @zentol question: by the fact that timeouts are always passed, but are never explicitly addressed in the method calls, I assume that they are handled by the underlying RPC machinery automatically? I switched the timeout type from Time to Duration - will it break the proper handling? I'm not entirely sure but would expect that RPC timeouts must be based on Flinks Time class. > This seems to be a "new" approach that was not previously used for the TaskExecutorGateway, also for similarly-scoped methods like requestThreadDump. Could you expand on the idea a bit? Should it be a single method interface just for the thread info collection purposes or do you want to use it to potentially group things. What do you think would be the appropriate name? Yes, we failed to do this in the past, and paid the price for it later on :( A SAM should suffice, for the name we could maybe use `ThreadDumpProvider`? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
