Say we have one worker with 4 CPU cores. How does parallelism configured in Dataflow worker machines? Do we parallelize beyond # of cores?
Where would this type of information be available?
可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效，请关闭广告屏蔽插件后再试):
问题:
回答1:
One worker thread is used per core, and each worker thread independently processes a chunk of the input space.
