I noticed that in my usage, a worker could hang forever after running the following statement, the timeout guard is not in effective, I did applied the changes and so far I didn't have this issue any more.
self.decoder_pipeline.end_request()
self.state = self.STATE_EOS_RECEIVED
I noticed that in my usage, a worker could hang forever after running the following statement, the timeout guard is not in effective, I did applied the changes and so far I didn't have this issue any more. self.decoder_pipeline.end_request() self.state = self.STATE_EOS_RECEIVED