hachikuji commented on a change in pull request #9590:
URL: https://github.com/apache/kafka/pull/9590#discussion_r537811235
##########
File path:
clients/src/main/java/org/apache/kafka/common/record/MemoryRecords.java
##########
@@ -198,7 +198,7 @@ private static FilterResult filterTo(TopicPartition
partition, Iterable<MutableR
if (!retainedRecords.isEmpty()) {
if (writeOriginalBatch) {
batch.writeTo(bufferOutputStream);
- filterResult.updateRetainedBatchMetadata(batch,
retainedRecords.size(), false);
+ filterResult.updateRetainedBatchMetadata(batch,
retainedRecords.get(0).offset(), retainedRecords.size(), false);
Review comment:
I don't think we have much choice given retention semantics. I think we
have to keep the invariant that the segment base offset is less than or equal
to the base offset of any batch contained in it. Note that these batches are
still returned in Fetch requests, so it seems reasonable to let the log start
offset reflect them.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
