You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kafka.apache.org by "Bruno Cadonna (Jira)" <ji...@apache.org> on 2019/11/01 15:32:00 UTC

[jira] [Created] (KAFKA-9129) Add Thread ID to the InternalProcessorContext

Bruno Cadonna created KAFKA-9129:
------------------------------------

             Summary: Add Thread ID to the InternalProcessorContext
                 Key: KAFKA-9129
                 URL: https://issues.apache.org/jira/browse/KAFKA-9129
             Project: Kafka
          Issue Type: Improvement
          Components: streams
            Reporter: Bruno Cadonna


When we added client metrics we had to move the {{StreamsMetricsImpl}} object to the client level. That means that now instead of having one {{StreamsMetricsImpl}} object per thread, we have now one per client. That also means that we cannot store the thread ID in the {{StreamsMetricsImpl}} anymore. Currently, we get the thread ID from {{Thread.currentThread().getName()}} when we need to create a sensor. However, that is not robust against code refactoring because we need to ensure that the thread that creates the sensor is also the one that records the metrics. To be more flexible, we should expose the ID of the thread that executes a processor in the {{InternalProcessorContext}} like it already exposes the task ID.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)