smjn commented on code in PR #18514:
URL: https://github.com/apache/kafka/pull/18514#discussion_r1914382884
##########
core/src/main/java/kafka/server/share/SharePartition.java:
##########
@@ -456,6 +456,17 @@ public CompletableFuture<Void> maybeInitialize() {
stateBatch.lastOffset(),
RecordState.forId(stateBatch.deliveryState()), stateBatch.deliveryCount(),
null);
cachedState.put(stateBatch.firstOffset(), inFlightBatch);
}
+
+ long initialOffset = startOffset;
Review Comment:
startOffset is 10 and is never updated in the loop -
for 2 batches (15, 20, 2, 1), (21, 25, 2, 2)
initial = 10
first batch
10 != 15 => if loop is executed newBatch = (10, 14, 0, 0), cachedState:{10
-> (10,14,0,0)}, initial = 21
second batch
10 != 21 => if loop is executed newBatch = (21, 20, 0, 0), cachedState:{10
-> (10,14,0,0), 21 -> (21,20,0,0)}, initial = 26
what am I missing?
This might work
```
soFar = startOffset
// assuming non-overlapping, sorted batches by first, last, deliveryCount,
deliveryState
for(Batch b: batches) {
// case 1:
// | <--- start offset
// ------- <--- interval
if soFar < b.firstOffset()
Batch newBatch = new Batch(start, b.firstOffset()-1, 0, 0)
cache.put(soFar, newBatch)
// case 2:
// | // firstOffset intersection (No new batch)
// --------
else
cache.put(soFar, batch)
// case 3:
// | // intersection somewhere in the middle (not
possible due to persister)
// --------
// case 4:
// | // intersection at last offset (not possible due to
persister)
// --------
// case 5:
// | // interval lastOffset < startOffset (not possible due
to persister)
// ------
soFar = batch.lastOffset() + 1
}
```
--
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: jira-unsubscr...@kafka.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
