showuon commented on code in PR #15474:
URL: https://github.com/apache/kafka/pull/15474#discussion_r1520857036
##########
clients/src/main/java/org/apache/kafka/common/record/MemoryRecordsBuilder.java:
##########
@@ -263,13 +262,8 @@ public RecordsInfo info() {
} else if (maxTimestamp == RecordBatch.NO_TIMESTAMP) {
return new RecordsInfo(RecordBatch.NO_TIMESTAMP, lastOffset);
} else {
- long shallowOffsetOfMaxTimestamp;
- // Use the last offset when dealing with record batches
- if (compressionType != CompressionType.NONE || magic >=
RecordBatch.MAGIC_VALUE_V2)
- shallowOffsetOfMaxTimestamp = lastOffset;
- else
- shallowOffsetOfMaxTimestamp = offsetOfMaxTimestamp;
- return new RecordsInfo(maxTimestamp, shallowOffsetOfMaxTimestamp);
+ // For create time, we always use offsetOfMaxTimestamp for the
correct time -> offset mapping
+ return new RecordsInfo(maxTimestamp, offsetOfMaxTimestamp);
Review Comment:
> This has the same issue. The semantic for MAX_TIMESTAMP is the first
offset with the max timestamp. So, if timestamp is
TimestampType.LOG_APPEND_TIME, we need to use the baseOffset, instead of
lastOffset.
Yes, updated.
> Also, could we remove the shallow part in
RecordsInfo.shallowOffsetOfMaxTimestamp?
I've added a comment in another PR:
https://github.com/apache/kafka/pull/15476#issuecomment-1990459827 . We can do
all the renaming there.
##########
clients/src/main/java/org/apache/kafka/common/record/MemoryRecordsBuilder.java:
##########
@@ -263,13 +262,8 @@ public RecordsInfo info() {
} else if (maxTimestamp == RecordBatch.NO_TIMESTAMP) {
return new RecordsInfo(RecordBatch.NO_TIMESTAMP, lastOffset);
} else {
- long shallowOffsetOfMaxTimestamp;
- // Use the last offset when dealing with record batches
- if (compressionType != CompressionType.NONE || magic >=
RecordBatch.MAGIC_VALUE_V2)
- shallowOffsetOfMaxTimestamp = lastOffset;
- else
- shallowOffsetOfMaxTimestamp = offsetOfMaxTimestamp;
- return new RecordsInfo(maxTimestamp, shallowOffsetOfMaxTimestamp);
+ // For create time, we always use offsetOfMaxTimestamp for the
correct time -> offset mapping
+ return new RecordsInfo(maxTimestamp, offsetOfMaxTimestamp);
Review Comment:
> This has the same issue. The semantic for MAX_TIMESTAMP is the first
offset with the max timestamp. So, if timestamp is
TimestampType.LOG_APPEND_TIME, we need to use the baseOffset, instead of
lastOffset.
Yes, updated.
> Also, could we remove the shallow part in
RecordsInfo.shallowOffsetOfMaxTimestamp?
I've added a comment in another PR:
https://github.com/apache/kafka/pull/15476#issuecomment-1990459827 . We can do
all the renaming there.
--
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]
