Github user tzulitai commented on a diff in the pull request:
https://github.com/apache/flink/pull/2687#discussion_r105688259
--- Diff:
flink-connectors/flink-connector-kafka-base/src/main/java/org/apache/flink/streaming/connectors/kafka/FlinkKafkaConsumerBase.java
---
@@ -558,11 +629,31 @@ protected static void
initializeSubscribedPartitionsToStartOffsets(
List<KafkaTopicPartition> kafkaTopicPartitions,
int indexOfThisSubtask,
int numParallelSubtasks,
- StartupMode startupMode) {
+ StartupMode startupMode,
+ Map<KafkaTopicPartition, Long> specificStartupOffsets) {
for (int i = 0; i < kafkaTopicPartitions.size(); i++) {
if (i % numParallelSubtasks == indexOfThisSubtask) {
-
subscribedPartitionsToStartOffsets.put(kafkaTopicPartitions.get(i),
startupMode.getStateSentinel());
+ if (startupMode !=
StartupMode.SPECIFIC_OFFSETS) {
+
subscribedPartitionsToStartOffsets.put(kafkaTopicPartitions.get(i),
startupMode.getStateSentinel());
+ } else {
+ if (specificStartupOffsets == null) {
+ throw new
IllegalArgumentException(
+ "Startup mode for the
consumer set to " + StartupMode.SPECIFIC_OFFSETS +
+ ", but no
specific offsets were specified");
+ }
+
+ KafkaTopicPartition partition =
kafkaTopicPartitions.get(i);
+
+ Long specificOffset =
specificStartupOffsets.get(partition);
+ if (specificOffset != null) {
+ // since the specified offsets
represent the next record to read, we subtract
+ // it by one so that the
initial state of the consumer will be correct
+
subscribedPartitionsToStartOffsets.put(partition, specificOffset - 1);
--- End diff --
That will simply be treated as an invalid offset (if the record with offset
0 doesn't exist; if it actually exists, I don't think there's much we can do
about it). So, again, the `auto.offset.reset` setting is automatically used by
the clients.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
