mimaison commented on code in PR #19285: URL: https://github.com/apache/kafka/pull/19285#discussion_r2029279945
########## storage/src/main/java/org/apache/kafka/server/purgatory/DelayedRemoteListOffsets.java: ########## @@ -0,0 +1,200 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.kafka.server.purgatory; + +import org.apache.kafka.common.TopicPartition; +import org.apache.kafka.common.errors.ApiException; +import org.apache.kafka.common.message.ListOffsetsResponseData; +import org.apache.kafka.common.protocol.Errors; +import org.apache.kafka.common.record.FileRecords; +import org.apache.kafka.common.requests.ListOffsetsResponse; +import org.apache.kafka.server.metrics.KafkaMetricsGroup; +import org.apache.kafka.storage.internals.log.OffsetResultHolder; + +import com.yammer.metrics.core.Meter; + +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import java.util.ArrayList; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.Optional; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.ExecutionException; +import java.util.concurrent.TimeUnit; +import java.util.concurrent.atomic.AtomicBoolean; +import java.util.function.Consumer; + +public class DelayedRemoteListOffsets extends DelayedOperation { + + private static final Logger LOG = LoggerFactory.getLogger(DelayedRemoteListOffsets.class); + + // For compatibility, metrics are defined to be under `kafka.server.DelayedRemoteListOffsetsMetrics` class + private static final KafkaMetricsGroup METRICS_GROUP = new KafkaMetricsGroup("kafka.server", "DelayedRemoteListOffsetsMetrics"); + static final Meter AGGREGATE_EXPIRATION_METER = METRICS_GROUP.newMeter("ExpiresPerSec", "requests", TimeUnit.SECONDS); + static final Map<TopicPartition, Meter> PARTITION_EXPIRATION_METERS = new ConcurrentHashMap<>(); + + private final int version; + private final Map<TopicPartition, ListOffsetsPartitionStatus> statusByPartition; + private final Consumer<TopicPartition> partitionOrException; + private final Consumer<List<ListOffsetsResponseData.ListOffsetsTopicResponse>> responseCallback; + + public DelayedRemoteListOffsets(long delayMs, + int version, + Map<TopicPartition, ListOffsetsPartitionStatus> statusByPartition, + Consumer<TopicPartition> partitionOrException, + Consumer<List<ListOffsetsResponseData.ListOffsetsTopicResponse>> responseCallback) { + super(delayMs); + this.version = version; + this.statusByPartition = statusByPartition; + this.partitionOrException = partitionOrException; + this.responseCallback = responseCallback; + // Mark the status as completed, if there is no async task to track. + // If there is a task to track, then build the response as REQUEST_TIMED_OUT by default. + statusByPartition.forEach((topicPartition, status) -> { + status.completed(status.futureHolderOpt().isEmpty()); + if (status.futureHolderOpt().isPresent()) { + status.responseOpt(Optional.of(buildErrorResponse(Errors.REQUEST_TIMED_OUT, topicPartition.partition()))); + } + LOG.trace("Initial partition status for {} is {}", topicPartition, status); + }); + } + + /** + * Call-back to execute when a delayed operation gets expired and hence forced to complete. + */ + @Override + public void onExpiration() { + statusByPartition.forEach((topicPartition, status) -> { + if (!status.completed()) { + LOG.debug("Expiring list offset request for partition {} with status {}", topicPartition, status); + status.futureHolderOpt().ifPresent(futureHolder -> futureHolder.jobFuture().cancel(true)); + recordExpiration(topicPartition); + } + }); + } + + /** + * Process for completing an operation; This function needs to be defined + * in subclasses and will be called exactly once in forceComplete() + */ + @Override + public void onComplete() { + Map<String, List<ListOffsetsResponseData.ListOffsetsPartitionResponse>> groupedByTopic = new HashMap<>(); + for (Map.Entry<TopicPartition, ListOffsetsPartitionStatus> entry : statusByPartition.entrySet()) { + List<ListOffsetsResponseData.ListOffsetsPartitionResponse> partitions = + groupedByTopic.computeIfAbsent(entry.getKey().topic(), k -> new ArrayList<>()); + if (entry.getValue().responseOpt().isPresent()) { + partitions.add(entry.getValue().responseOpt().get()); + } + } + List<ListOffsetsResponseData.ListOffsetsTopicResponse> response = new ArrayList<>(); + for (Map.Entry<String, List<ListOffsetsResponseData.ListOffsetsPartitionResponse>> entry : groupedByTopic.entrySet()) { + response.add(new ListOffsetsResponseData.ListOffsetsTopicResponse() + .setName(entry.getKey()) + .setPartitions(entry.getValue())); + } + responseCallback.accept(response); Review Comment: Good idea, this can save us iterating over the responses twice. ########## storage/src/main/java/org/apache/kafka/server/purgatory/DelayedRemoteListOffsets.java: ########## @@ -0,0 +1,200 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.kafka.server.purgatory; + +import org.apache.kafka.common.TopicPartition; +import org.apache.kafka.common.errors.ApiException; +import org.apache.kafka.common.message.ListOffsetsResponseData; +import org.apache.kafka.common.protocol.Errors; +import org.apache.kafka.common.record.FileRecords; +import org.apache.kafka.common.requests.ListOffsetsResponse; +import org.apache.kafka.server.metrics.KafkaMetricsGroup; +import org.apache.kafka.storage.internals.log.OffsetResultHolder; + +import com.yammer.metrics.core.Meter; + +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import java.util.ArrayList; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.Optional; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.ExecutionException; +import java.util.concurrent.TimeUnit; +import java.util.concurrent.atomic.AtomicBoolean; +import java.util.function.Consumer; + +public class DelayedRemoteListOffsets extends DelayedOperation { + + private static final Logger LOG = LoggerFactory.getLogger(DelayedRemoteListOffsets.class); + + // For compatibility, metrics are defined to be under `kafka.server.DelayedRemoteListOffsetsMetrics` class + private static final KafkaMetricsGroup METRICS_GROUP = new KafkaMetricsGroup("kafka.server", "DelayedRemoteListOffsetsMetrics"); + static final Meter AGGREGATE_EXPIRATION_METER = METRICS_GROUP.newMeter("ExpiresPerSec", "requests", TimeUnit.SECONDS); + static final Map<TopicPartition, Meter> PARTITION_EXPIRATION_METERS = new ConcurrentHashMap<>(); + + private final int version; + private final Map<TopicPartition, ListOffsetsPartitionStatus> statusByPartition; + private final Consumer<TopicPartition> partitionOrException; + private final Consumer<List<ListOffsetsResponseData.ListOffsetsTopicResponse>> responseCallback; + + public DelayedRemoteListOffsets(long delayMs, + int version, + Map<TopicPartition, ListOffsetsPartitionStatus> statusByPartition, + Consumer<TopicPartition> partitionOrException, + Consumer<List<ListOffsetsResponseData.ListOffsetsTopicResponse>> responseCallback) { + super(delayMs); + this.version = version; + this.statusByPartition = statusByPartition; + this.partitionOrException = partitionOrException; + this.responseCallback = responseCallback; + // Mark the status as completed, if there is no async task to track. + // If there is a task to track, then build the response as REQUEST_TIMED_OUT by default. + statusByPartition.forEach((topicPartition, status) -> { + status.completed(status.futureHolderOpt().isEmpty()); + if (status.futureHolderOpt().isPresent()) { + status.responseOpt(Optional.of(buildErrorResponse(Errors.REQUEST_TIMED_OUT, topicPartition.partition()))); + } + LOG.trace("Initial partition status for {} is {}", topicPartition, status); + }); + } + + /** + * Call-back to execute when a delayed operation gets expired and hence forced to complete. + */ + @Override + public void onExpiration() { + statusByPartition.forEach((topicPartition, status) -> { + if (!status.completed()) { + LOG.debug("Expiring list offset request for partition {} with status {}", topicPartition, status); + status.futureHolderOpt().ifPresent(futureHolder -> futureHolder.jobFuture().cancel(true)); + recordExpiration(topicPartition); + } + }); + } + + /** + * Process for completing an operation; This function needs to be defined + * in subclasses and will be called exactly once in forceComplete() + */ + @Override + public void onComplete() { + Map<String, List<ListOffsetsResponseData.ListOffsetsPartitionResponse>> groupedByTopic = new HashMap<>(); + for (Map.Entry<TopicPartition, ListOffsetsPartitionStatus> entry : statusByPartition.entrySet()) { + List<ListOffsetsResponseData.ListOffsetsPartitionResponse> partitions = + groupedByTopic.computeIfAbsent(entry.getKey().topic(), k -> new ArrayList<>()); + if (entry.getValue().responseOpt().isPresent()) { + partitions.add(entry.getValue().responseOpt().get()); + } + } + List<ListOffsetsResponseData.ListOffsetsTopicResponse> response = new ArrayList<>(); + for (Map.Entry<String, List<ListOffsetsResponseData.ListOffsetsPartitionResponse>> entry : groupedByTopic.entrySet()) { + response.add(new ListOffsetsResponseData.ListOffsetsTopicResponse() + .setName(entry.getKey()) + .setPartitions(entry.getValue())); + } + responseCallback.accept(response); + } + + /** + * Try to complete the delayed operation by first checking if the operation + * can be completed by now. If yes execute the completion logic by calling + * forceComplete() and return true iff forceComplete returns true; otherwise return false + */ + @Override + public boolean tryComplete() { + AtomicBoolean completable = new AtomicBoolean(true); + statusByPartition.forEach((partition, status) -> { + if (!status.completed()) { + try { + partitionOrException.accept(partition); + } catch (ApiException e) { + status.futureHolderOpt().ifPresent(futureHolder -> { + futureHolder.jobFuture().cancel(false); + futureHolder.taskFuture().complete(new OffsetResultHolder.FileRecordsOrError(Optional.of(e), Optional.empty())); + }); + } + + status.futureHolderOpt().ifPresent(futureHolder -> { + if (futureHolder.taskFuture().isDone()) { + ListOffsetsResponseData.ListOffsetsPartitionResponse response; + try { + OffsetResultHolder.FileRecordsOrError taskFuture = futureHolder.taskFuture().get(); + if (taskFuture.hasException()) { + response = buildErrorResponse(Errors.forException(taskFuture.exception().get()), partition.partition()); + } else if (!taskFuture.hasTimestampAndOffset()) { + Errors error = status.maybeOffsetsError() + .map(e -> version >= 5 ? Errors.forException(e) : Errors.LEADER_NOT_AVAILABLE) + .orElse(Errors.NONE); + response = buildErrorResponse(error, partition.partition()); + } else { + ListOffsetsResponseData.ListOffsetsPartitionResponse partitionResponse = buildErrorResponse(Errors.NONE, partition.partition()); + FileRecords.TimestampAndOffset found = taskFuture.timestampAndOffset().get(); + if (status.lastFetchableOffset().isPresent() && found.offset >= status.lastFetchableOffset().get()) { + if (status.maybeOffsetsError().isPresent()) { + Errors error = version >= 5 ? Errors.forException(status.maybeOffsetsError().get()) : Errors.LEADER_NOT_AVAILABLE; + partitionResponse.setErrorCode(error.code()); + } + } else { + partitionResponse = new ListOffsetsResponseData.ListOffsetsPartitionResponse() + .setPartitionIndex(partition.partition()) + .setErrorCode(Errors.NONE.code()) + .setTimestamp(found.timestamp) + .setOffset(found.offset); + + if (found.leaderEpoch.isPresent() && version >= 4) { + partitionResponse.setLeaderEpoch(found.leaderEpoch.get()); + } + } + response = partitionResponse; + } + } catch (InterruptedException | ExecutionException e) { + response = buildErrorResponse(Errors.forException(e), partition.partition()); + } + status.responseOpt(Optional.of(response)); + status.completed(true); + } + completable.set(completable.get() && futureHolder.taskFuture().isDone()); + }); + } + }); + if (completable.get()) { + return forceComplete(); + } else { + return false; + } + } + + private ListOffsetsResponseData.ListOffsetsPartitionResponse buildErrorResponse(Errors e, int partitionIndex) { + return new ListOffsetsResponseData.ListOffsetsPartitionResponse() + .setPartitionIndex(partitionIndex) + .setErrorCode(e.code()) + .setTimestamp(ListOffsetsResponse.UNKNOWN_TIMESTAMP) + .setOffset(ListOffsetsResponse.UNKNOWN_OFFSET); + } + + private static void recordExpiration(TopicPartition partition) { + AGGREGATE_EXPIRATION_METER.mark(); + PARTITION_EXPIRATION_METERS.computeIfAbsent(partition, tp -> METRICS_GROUP.newMeter("ExpiresPerSec", + "requests", + TimeUnit.SECONDS, + Map.of("topic", tp.topic(), "partition", String.valueOf(tp.partition())))).mark(); Review Comment: Done -- 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
