mridulm commented on code in PR #37638: URL: https://github.com/apache/spark/pull/37638#discussion_r1003564958
########## common/network-shuffle/src/main/java/org/apache/spark/network/shuffle/RemoteBlockPushResolver.java: ########## @@ -1904,4 +1941,42 @@ long getPos() { return pos; } } + + /** + * A class that wraps all the push-based shuffle service metrics. + */ + static class PushMergeMetrics implements MetricSet { + // couldNotFindOpportunityResponses tracks how many times a shuffle block collided because + // of another block for the same reduce partition was being written + static final String NO_OPPORTUNITY_RESPONSES_METRIC = "couldNotFindOpportunityResponses"; + // tooLateResponses tracks how many times a shuffle block push request is too late + static final String TOO_LATE_RESPONSES_METRIC = "tooLateResponses"; + // pushedBytesWritten tracks the length of the pushed block data written to file in bytes + static final String PUSHED_BYTES_WRITTEN_METRIC = "pushedBytesWritten"; + // cachedBlocksBytes tracks the size of the current deferred block parts buffered in memory. + static final String CACHED_BLOCKS_BYTES_METRIC = "cachedBlocksBytes"; Review Comment: There is a correlation between number of deferred blocks and amount of bytes deferred - but they are not necessarily the same, and also are influenced by the spark application configuration. -- 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: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org