yaoliclshlmch commented on code in PR #18761:
URL: https://github.com/apache/hudi/pull/18761#discussion_r3277833496
##########
hudi-flink-datasource/hudi-flink/src/main/java/org/apache/hudi/metrics/FlinkIndexBackendMetrics.java:
##########
@@ -63,6 +76,22 @@ public void registerMetrics() {
metricGroup.histogram("remoteIndexLookupLatency",
remoteIndexLookupLatency);
metricGroup.histogram("localLookupKeysNum", localLookupKeysNum);
metricGroup.histogram("remoteLookupKeysNum", remoteLookupKeysNum);
+ metricGroup.counter(BUCKET_ASSIGN_MINIBATCH_CACHE_HIT_COUNT,
cacheHitCount);
+ metricGroup.counter(BUCKET_ASSIGN_MINIBATCH_CACHE_MISS_COUNT,
cacheMissCount);
Review Comment:
localLookupKeysNum and remoteLookupKeysNum are histogram, which has a 100s
aggregation, if we want the counter data point, we should keep these 2. we can
add the hit ratio gauge to get the ratio per mini batch., but keeping the cache
hit / miss count can give us the flexibility of querying the ratio at any time
range. wdyt?
--
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]
