hachikuji commented on a change in pull request #10909:
URL: https://github.com/apache/kafka/pull/10909#discussion_r679391787
##########
File path: raft/src/main/java/org/apache/kafka/raft/ReplicatedCounter.java
##########
@@ -61,10 +61,12 @@ public synchronized void increment() {
int epoch = claimedEpoch.getAsInt();
uncommitted += 1;
- Long offset = client.scheduleAppend(epoch, singletonList(uncommitted));
- if (offset != null) {
+ try {
+ long offset = client.scheduleAppend(epoch,
singletonList(uncommitted));
log.debug("Scheduled append of record {} with epoch {} at offset
{}",
uncommitted, epoch, offset);
+ } catch (IllegalStateException e) {
Review comment:
How about we catch `Exception` generally?
##########
File path:
raft/src/main/java/org/apache/kafka/raft/internals/BatchAccumulator.java
##########
@@ -148,7 +150,7 @@ private Long append(int epoch, List<T> records, boolean
isAtomic) {
}
if (batch == null) {
- return null;
+ throw new IllegalStateException("Append failed because we
failed to allocate memory to write the batch");
Review comment:
I wonder if it is worthwhile introducing some custom raft exception
types for these two cases. At a minimum, that would let us improve the
assertions. Maybe `NotLeaderException` and `BufferAllocationException` or
something like that?
--
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]
