cchung100m commented on code in PR #2600:
URL: https://github.com/apache/uniffle/pull/2600#discussion_r2326565975
##########
client-spark/common/src/main/java/org/apache/spark/shuffle/reader/RssShuffleDataIterator.java:
##########
@@ -117,17 +117,20 @@ public boolean hasNext() {
// If ShuffleServer delete
ByteBuffer rawData = rawBlock != null ? rawBlock.getByteBuffer() : null;
- long fetchDuration = System.currentTimeMillis() - startFetch;
- shuffleReadMetrics.incFetchWaitTime(fetchDuration);
+ long readDuration = System.currentTimeMillis() - startFetch;
if (rawData != null) {
uncompress(rawBlock, rawData);
// create new iterator for shuffle data
long startSerialization = System.currentTimeMillis();
recordsIterator = createKVIterator(uncompressedData);
long serializationDuration = System.currentTimeMillis() -
startSerialization;
- readTime += fetchDuration;
+ long fetchDuration = System.currentTimeMillis() - startFetch;
+ shuffleReadMetrics.incFetchWaitTime(fetchDuration);
+ readTime += readDuration;
serializeTime += serializationDuration;
Review Comment:
Hi @zuston
Thank you for getting back to me so quickly and for reminding me. I have
updated the part you mentioned.
Our new implementation does not seem to fit the original test case because
it takes the `fetchDuration`(network transmission) as an assumption. Do you
think we should change the test case here?
https://github.com/apache/uniffle/blob/1e48bc673d1c0ee41f889a0de6192b0fab131467/client-spark/common/src/test/java/org/apache/spark/shuffle/reader/RssShuffleDataIteratorTest.java#L337
--
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]
