You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2022/06/15 19:46:35 UTC
[GitHub] [flink] syhily opened a new pull request, #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
syhily opened a new pull request, #19974:
URL: https://github.com/apache/flink/pull/19974
## What is the purpose of the change
This pull request moves record deserializing in PulsarSource from SplitFetcher to RecordEmitter to support object-reusing deserializer.
## Brief change log
Move record deserializing from PulsarPartitionSplitFetcher to PulsarRecordEmitter
## Verifying this change
This change is already covered by existing tests, such as:
- PulsarOrderedPartitionSplitReaderTest
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): (no)
- The public API, i.e., is any changed class annotated with `@Public(Evolving)`: (no)
- The serializers: (no)
- The runtime per-record code paths (performance sensitive): (no)
- Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: (no)
- The S3 file system connector: (no)
## Documentation
- Does this pull request introduce a new feature? (no)
- If yes, how is the feature documented? (not applicable)
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] reswqa commented on a diff in pull request #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
Posted by GitBox <gi...@apache.org>.
reswqa commented on code in PR #19974:
URL: https://github.com/apache/flink/pull/19974#discussion_r899950625
##########
flink-connectors/flink-connector-pulsar/src/main/java/org/apache/flink/connector/pulsar/source/reader/emitter/PulsarRecordEmitter.java:
##########
@@ -20,26 +20,69 @@
import org.apache.flink.api.connector.source.SourceOutput;
import org.apache.flink.connector.base.source.reader.RecordEmitter;
-import org.apache.flink.connector.pulsar.source.reader.message.PulsarMessage;
+import org.apache.flink.connector.pulsar.source.reader.deserializer.PulsarDeserializationSchema;
import org.apache.flink.connector.pulsar.source.reader.source.PulsarOrderedSourceReader;
import org.apache.flink.connector.pulsar.source.reader.source.PulsarUnorderedSourceReader;
import org.apache.flink.connector.pulsar.source.split.PulsarPartitionSplitState;
+import org.apache.flink.util.Collector;
+
+import org.apache.pulsar.client.api.Message;
/**
* The {@link RecordEmitter} implementation for both {@link PulsarOrderedSourceReader} and {@link
* PulsarUnorderedSourceReader}. We would always update the last consumed message id in this
* emitter.
*/
public class PulsarRecordEmitter<T>
- implements RecordEmitter<PulsarMessage<T>, T, PulsarPartitionSplitState> {
+ implements RecordEmitter<Message<byte[]>, T, PulsarPartitionSplitState> {
+
+ private final PulsarDeserializationSchema<T> deserializationSchema;
+ private final SourceOutputWrapper<T> sourceOutputWrapper = new SourceOutputWrapper<>();
+
+ public PulsarRecordEmitter(PulsarDeserializationSchema<T> deserializationSchema) {
+ this.deserializationSchema = deserializationSchema;
+ }
@Override
public void emitRecord(
- PulsarMessage<T> element, SourceOutput<T> output, PulsarPartitionSplitState splitState)
+ Message<byte[]> element, SourceOutput<T> output, PulsarPartitionSplitState splitState)
throws Exception {
- // Sink the record to source output.
- output.collect(element.getValue(), element.getEventTime());
- // Update the split state.
- splitState.setLatestConsumedId(element.getId());
+ // Update the source output.
+ sourceOutputWrapper.setSourceOutput(output);
+ sourceOutputWrapper.setTimestamp(element);
+
+ deserializationSchema.deserialize(element, sourceOutputWrapper);
+ splitState.setLatestConsumedId(element.getMessageId());
+ }
+
+ private static class SourceOutputWrapper<T> implements Collector<T> {
+ private SourceOutput<T> sourceOutput;
+ private long timestamp;
+
+ @Override
+ public void collect(T record) {
+ if (timestamp > 0) {
+ sourceOutput.collect(record, timestamp);
+ } else {
+ sourceOutput.collect(record);
+ }
+ }
+
+ @Override
+ public void close() {
+ // Nothing to do here.
+ }
+
+ private void setSourceOutput(SourceOutput<T> sourceOutput) {
+ this.sourceOutput = sourceOutput;
+ }
+
+ /**
+ * Get the event timestamp from Pulsar. Zero means there is no event time. See {@link
+ * Message#getEventTime()} to get the reason why it returns zero.
+ */
+ private void setTimestamp(Message<?> message) {
+ this.timestamp = message.getEventTime();
Review Comment:
Why use Message as parameter instead of passing timestamp directly
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] flinkbot commented on pull request #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
Posted by GitBox <gi...@apache.org>.
flinkbot commented on PR #19974:
URL: https://github.com/apache/flink/pull/19974#issuecomment-1156861952
<!--
Meta data
{
"version" : 1,
"metaDataEntries" : [ {
"hash" : "2dc70b01fc3387cab385936f03069075471aecac",
"status" : "UNKNOWN",
"url" : "TBD",
"triggerID" : "2dc70b01fc3387cab385936f03069075471aecac",
"triggerType" : "PUSH"
} ]
}-->
## CI report:
* 2dc70b01fc3387cab385936f03069075471aecac UNKNOWN
<details>
<summary>Bot commands</summary>
The @flinkbot bot supports the following commands:
- `@flinkbot run azure` re-run the last Azure build
</details>
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] imaffe commented on pull request #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
Posted by GitBox <gi...@apache.org>.
imaffe commented on PR #19974:
URL: https://github.com/apache/flink/pull/19974#issuecomment-1157166125
I'm not very familiar with the object-reuse deserialization schema here. Could you describe briefly why we introduce this change ?
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] syhily closed pull request #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
Posted by GitBox <gi...@apache.org>.
syhily closed pull request #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
URL: https://github.com/apache/flink/pull/19974
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] syhily commented on a diff in pull request #19974: [FLINK-28083][Connector/Pulsar] Object-reusing for Pulsar source
Posted by GitBox <gi...@apache.org>.
syhily commented on code in PR #19974:
URL: https://github.com/apache/flink/pull/19974#discussion_r901427307
##########
flink-connectors/flink-connector-pulsar/src/main/java/org/apache/flink/connector/pulsar/source/reader/emitter/PulsarRecordEmitter.java:
##########
@@ -20,26 +20,69 @@
import org.apache.flink.api.connector.source.SourceOutput;
import org.apache.flink.connector.base.source.reader.RecordEmitter;
-import org.apache.flink.connector.pulsar.source.reader.message.PulsarMessage;
+import org.apache.flink.connector.pulsar.source.reader.deserializer.PulsarDeserializationSchema;
import org.apache.flink.connector.pulsar.source.reader.source.PulsarOrderedSourceReader;
import org.apache.flink.connector.pulsar.source.reader.source.PulsarUnorderedSourceReader;
import org.apache.flink.connector.pulsar.source.split.PulsarPartitionSplitState;
+import org.apache.flink.util.Collector;
+
+import org.apache.pulsar.client.api.Message;
/**
* The {@link RecordEmitter} implementation for both {@link PulsarOrderedSourceReader} and {@link
* PulsarUnorderedSourceReader}. We would always update the last consumed message id in this
* emitter.
*/
public class PulsarRecordEmitter<T>
- implements RecordEmitter<PulsarMessage<T>, T, PulsarPartitionSplitState> {
+ implements RecordEmitter<Message<byte[]>, T, PulsarPartitionSplitState> {
+
+ private final PulsarDeserializationSchema<T> deserializationSchema;
+ private final SourceOutputWrapper<T> sourceOutputWrapper = new SourceOutputWrapper<>();
+
+ public PulsarRecordEmitter(PulsarDeserializationSchema<T> deserializationSchema) {
+ this.deserializationSchema = deserializationSchema;
+ }
@Override
public void emitRecord(
- PulsarMessage<T> element, SourceOutput<T> output, PulsarPartitionSplitState splitState)
+ Message<byte[]> element, SourceOutput<T> output, PulsarPartitionSplitState splitState)
throws Exception {
- // Sink the record to source output.
- output.collect(element.getValue(), element.getEventTime());
- // Update the split state.
- splitState.setLatestConsumedId(element.getId());
+ // Update the source output.
+ sourceOutputWrapper.setSourceOutput(output);
+ sourceOutputWrapper.setTimestamp(element);
+
+ deserializationSchema.deserialize(element, sourceOutputWrapper);
+ splitState.setLatestConsumedId(element.getMessageId());
+ }
+
+ private static class SourceOutputWrapper<T> implements Collector<T> {
+ private SourceOutput<T> sourceOutput;
+ private long timestamp;
+
+ @Override
+ public void collect(T record) {
+ if (timestamp > 0) {
+ sourceOutput.collect(record, timestamp);
+ } else {
+ sourceOutput.collect(record);
+ }
+ }
+
+ @Override
+ public void close() {
+ // Nothing to do here.
+ }
+
+ private void setSourceOutput(SourceOutput<T> sourceOutput) {
+ this.sourceOutput = sourceOutput;
+ }
+
+ /**
+ * Get the event timestamp from Pulsar. Zero means there is no event time. See {@link
+ * Message#getEventTime()} to get the reason why it returns zero.
+ */
+ private void setTimestamp(Message<?> message) {
+ this.timestamp = message.getEventTime();
Review Comment:
Just to avoid the same code as Kafka.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org