stevenschlansker opened a new pull request, #20328: URL: https://github.com/apache/kafka/pull/20328
While diagnosing an out of memory condition, our profiler alerted us that tens of megabytes of our heap was used to store redundant String objects. In particular, the strings `thread_id`, `rocksdb_state_id`, `task_id`, and `org.apache.kafka.consumer.fetch.manager.preferred.read.replica` alone take 12MB. All of these objects are strongly held by MetricKey or MetricName instances. Interning a string is a somewhat expensive operation, however, metric names are generally created once at initialization and then not again dynamically (excepting some per-connection tracking) -- 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. To unsubscribe, e-mail: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
