gortiz opened a new pull request, #16576: URL: https://github.com/apache/pinot/pull/16576
This PR includes two new stats for multi-stage operators: 1. `allocatedMemoryBytes`: The number of bytes allocated by this operator and its children on the same stage. It uses the same mechanism as SSE to populate `threadMemAllocatedBytes` (which is also used by OOM killing). This means it is only collected if `pinot.server.instance.enableThreadAllocatedBytesMeasurement` and/or `pinot.broker.instance.enableThreadAllocatedBytesMeasurement` is enabled. 2. `gcTimeMs`: The number of milliseconds spent on GC while executing this operator. It is not clear to me how expensive it is to call this method, so give this should be called once per nextBlock per operator, it is disabled by default. It can be enabled using the new query option `collectGCStats`. These stats will be handy for users to find out the most expensive operator on a query, but also for Pinot devs to find operators that are not very efficient. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
