[GitHub] [kafka] showuon commented on a diff in pull request #13760: KAFKA-8982: Add retry of fetching metadata to Admin.deleteRecords
showuon commented on code in PR #13760: URL: https://github.com/apache/kafka/pull/13760#discussion_r1238266317 ## clients/src/main/java/org/apache/kafka/clients/admin/internals/DeleteRecordsHandler.java: ## @@ -0,0 +1,174 @@ +/* + * 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.clients.admin.internals; + +import java.util.ArrayList; +import java.util.Collection; +import java.util.HashMap; +import java.util.HashSet; +import java.util.List; +import java.util.Map; +import java.util.Set; +import org.apache.kafka.clients.admin.DeletedRecords; +import org.apache.kafka.clients.admin.RecordsToDelete; +import org.apache.kafka.clients.admin.internals.AdminApiFuture.SimpleAdminApiFuture; +import org.apache.kafka.clients.admin.internals.AdminApiHandler.Batched; +import org.apache.kafka.common.Node; +import org.apache.kafka.common.TopicPartition; +import org.apache.kafka.common.errors.ApiException; +import org.apache.kafka.common.errors.InvalidMetadataException; +import org.apache.kafka.common.errors.RetriableException; +import org.apache.kafka.common.errors.TopicAuthorizationException; +import org.apache.kafka.common.message.DeleteRecordsRequestData; +import org.apache.kafka.common.message.DeleteRecordsResponseData; +import org.apache.kafka.common.protocol.Errors; +import org.apache.kafka.common.requests.AbstractResponse; +import org.apache.kafka.common.requests.DeleteRecordsRequest; +import org.apache.kafka.common.requests.DeleteRecordsResponse; +import org.apache.kafka.common.utils.LogContext; +import org.slf4j.Logger; + +public final class DeleteRecordsHandler extends Batched { + +private final Map recordsToDelete; +private final Logger log; +private final AdminApiLookupStrategy lookupStrategy; + +public DeleteRecordsHandler( +Map recordsToDelete, +LogContext logContext +) { +this.recordsToDelete = recordsToDelete; +this.log = logContext.logger(DeleteRecordsHandler.class); +this.lookupStrategy = new PartitionLeaderStrategy(logContext); +} + +@Override +public String apiName() { +return "deleteRecords"; +} + +@Override +public AdminApiLookupStrategy lookupStrategy() { +return this.lookupStrategy; +} + +public static SimpleAdminApiFuture newFuture( +Collection topicPartitions +) { +return AdminApiFuture.forKeys(new HashSet<>(topicPartitions)); +} + +@Override +public DeleteRecordsRequest.Builder buildBatchedRequest(int brokerId, Set keys) { +Map deletionsForTopic = new HashMap<>(); +for (Map.Entry entry: recordsToDelete.entrySet()) { +TopicPartition topicPartition = entry.getKey(); +DeleteRecordsRequestData.DeleteRecordsTopic deleteRecords = deletionsForTopic.computeIfAbsent( +topicPartition.topic(), +key -> new DeleteRecordsRequestData.DeleteRecordsTopic().setName(topicPartition.topic()) +); +deleteRecords.partitions().add(new DeleteRecordsRequestData.DeleteRecordsPartition() +.setPartitionIndex(topicPartition.partition()) +.setOffset(entry.getValue().beforeOffset())); +} + +DeleteRecordsRequestData data = new DeleteRecordsRequestData() +.setTopics(new ArrayList<>(deletionsForTopic.values())); Review Comment: No, I don't think they are the same thing. The invokeDriver timeout is set in the client side for client timeout the request. But the `timeoutMs` field in `DeleteRecordsRequest` is to send to the server side to notify the server about the timeout. REF: https://github.com/apache/kafka/blob/trunk/clients/src/main/resources/common/message/DeleteRecordsRequest.json#L39 -- 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
[GitHub] [kafka] showuon commented on a diff in pull request #13760: KAFKA-8982: Add retry of fetching metadata to Admin.deleteRecords
showuon commented on code in PR #13760: URL: https://github.com/apache/kafka/pull/13760#discussion_r1238257723 ## clients/src/test/java/org/apache/kafka/clients/admin/KafkaAdminClientTest.java: ## @@ -2359,31 +2334,22 @@ public void testDeleteRecords() throws Exception { assertTrue(e0.getCause() instanceof OffsetOutOfRangeException); } -// "leader not available" failure on metadata request for partition 2 +// not authorized to delete records for partition 2 KafkaFuture myTopicPartition2Result = values.get(myTopicPartition2); try { myTopicPartition2Result.get(); fail("get() should throw ExecutionException"); } catch (ExecutionException e1) { -assertTrue(e1.getCause() instanceof LeaderNotAvailableException); +assertTrue(e1.getCause() instanceof TopicAuthorizationException); } Review Comment: Ah, I see. Thanks for the explanation. -- 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
[GitHub] [kafka] showuon commented on a diff in pull request #13760: KAFKA-8982: Add retry of fetching metadata to Admin.deleteRecords
showuon commented on code in PR #13760: URL: https://github.com/apache/kafka/pull/13760#discussion_r1229107591 ## clients/src/test/java/org/apache/kafka/clients/admin/KafkaAdminClientTest.java: ## @@ -2359,31 +2334,22 @@ public void testDeleteRecords() throws Exception { assertTrue(e0.getCause() instanceof OffsetOutOfRangeException); } -// "leader not available" failure on metadata request for partition 2 +// not authorized to delete records for partition 2 KafkaFuture myTopicPartition2Result = values.get(myTopicPartition2); try { myTopicPartition2Result.get(); fail("get() should throw ExecutionException"); } catch (ExecutionException e1) { -assertTrue(e1.getCause() instanceof LeaderNotAvailableException); +assertTrue(e1.getCause() instanceof TopicAuthorizationException); } Review Comment: This can be replaced with `assertThrows(TopicAuthorizationException.class, () ->myTopicPartition2Result.get(), "get() should throw ExecutionException" )` Also, Should we change the comment into `TopicAuthorizationException`? It doesn't throw `ExecutionException` at all. Same comment applies to other places. ## clients/src/test/java/org/apache/kafka/clients/admin/internals/DeleteRecordsHandlerTest.java: ## @@ -0,0 +1,209 @@ +/* + * 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.clients.admin.internals; + +import java.util.ArrayList; +import java.util.HashMap; +import java.util.HashSet; +import java.util.List; +import java.util.Map; +import java.util.Set; +import org.apache.kafka.clients.admin.DeletedRecords; +import org.apache.kafka.clients.admin.RecordsToDelete; +import org.apache.kafka.common.Node; +import org.apache.kafka.common.TopicPartition; +import org.apache.kafka.common.message.DeleteRecordsRequestData; +import org.apache.kafka.common.message.DeleteRecordsResponseData; +import org.apache.kafka.common.protocol.Errors; +import org.apache.kafka.common.requests.DeleteRecordsRequest; +import org.apache.kafka.common.requests.DeleteRecordsResponse; +import org.apache.kafka.common.utils.LogContext; +import org.junit.jupiter.api.Test; + +import static java.util.Collections.emptyList; +import static java.util.Collections.emptyMap; +import static java.util.Collections.emptySet; +import static java.util.Collections.singleton; + +import static org.apache.kafka.common.utils.Utils.mkSet; +import static org.junit.jupiter.api.Assertions.assertEquals; +import static org.junit.jupiter.api.Assertions.assertTrue; + +public class DeleteRecordsHandlerTest { +private final LogContext logContext = new LogContext(); +private final TopicPartition t0p0 = new TopicPartition("t0", 0); +private final TopicPartition t0p1 = new TopicPartition("t0", 1); +private final Node node = new Node(1, "host", 1234); +private final Map recordsToDelete = new HashMap() { +{ +put(t0p0, RecordsToDelete.beforeOffset(10L)); +put(t0p1, RecordsToDelete.beforeOffset(10L)); +} +}; + +@Test +public void testBuildRequestSimple() { +DeleteRecordsHandler handler = new DeleteRecordsHandler(recordsToDelete, logContext); +DeleteRecordsRequest request = handler.buildBatchedRequest(node.id(), mkSet(t0p0, t0p1)).build(); +List topicPartitions = request.data().topics(); +assertEquals(1, topicPartitions.size()); +DeleteRecordsRequestData.DeleteRecordsTopic topic = topicPartitions.get(0); +assertEquals(2, topic.partitions().size()); +} + +@Test +public void testHandleSuccessfulResponse() { +AdminApiHandler.ApiResult result = +handleResponse(createResponse(emptyMap(), recordsToDelete.keySet())); +assertResult(result, recordsToDelete.keySet(), emptyMap(), emptyList(), emptySet()); +} + +@Test +public void testHandleRetriablePartitionTimeoutResponse() { +TopicPartition errorPartition = t0p0; +Map errorsByPartition = new HashMap<>(); +errorsByPartition.put(errorPartition,
[GitHub] [kafka] showuon commented on a diff in pull request #13760: KAFKA-8982: Add retry of fetching metadata to Admin.deleteRecords
showuon commented on code in PR #13760: URL: https://github.com/apache/kafka/pull/13760#discussion_r1228079251 ## clients/src/main/java/org/apache/kafka/clients/admin/KafkaAdminClient.java: ## @@ -2925,123 +2919,11 @@ void handleFailure(Throwable throwable) { @Override public DeleteRecordsResult deleteRecords(final Map recordsToDelete, final DeleteRecordsOptions options) { +AdminApiFuture.SimpleAdminApiFuture future = DeleteRecordsHandler.newFuture(recordsToDelete.keySet()); Review Comment: nit: `AdminApiFuture.SimpleAdminApiFuture` -> `SimpleAdminApiFuture` should work. ## clients/src/main/java/org/apache/kafka/clients/admin/internals/DeleteRecordsHandler.java: ## @@ -0,0 +1,174 @@ +/* + * 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.clients.admin.internals; + +import java.util.ArrayList; +import java.util.Collection; +import java.util.HashMap; +import java.util.HashSet; +import java.util.List; +import java.util.Map; +import java.util.Set; +import org.apache.kafka.clients.admin.DeletedRecords; +import org.apache.kafka.clients.admin.RecordsToDelete; +import org.apache.kafka.clients.admin.internals.AdminApiFuture.SimpleAdminApiFuture; +import org.apache.kafka.clients.admin.internals.AdminApiHandler.Batched; +import org.apache.kafka.common.Node; +import org.apache.kafka.common.TopicPartition; +import org.apache.kafka.common.errors.ApiException; +import org.apache.kafka.common.errors.InvalidMetadataException; +import org.apache.kafka.common.errors.RetriableException; +import org.apache.kafka.common.errors.TopicAuthorizationException; +import org.apache.kafka.common.message.DeleteRecordsRequestData; +import org.apache.kafka.common.message.DeleteRecordsResponseData; +import org.apache.kafka.common.protocol.Errors; +import org.apache.kafka.common.requests.AbstractResponse; +import org.apache.kafka.common.requests.DeleteRecordsRequest; +import org.apache.kafka.common.requests.DeleteRecordsResponse; +import org.apache.kafka.common.utils.LogContext; +import org.slf4j.Logger; + +public final class DeleteRecordsHandler extends Batched { + +private final Map recordsToDelete; +private final Logger log; +private final AdminApiLookupStrategy lookupStrategy; + +public DeleteRecordsHandler( +Map recordsToDelete, +LogContext logContext +) { +this.recordsToDelete = recordsToDelete; +this.log = logContext.logger(DeleteRecordsHandler.class); +this.lookupStrategy = new PartitionLeaderStrategy(logContext); +} + +@Override +public String apiName() { +return "deleteRecords"; +} + +@Override +public AdminApiLookupStrategy lookupStrategy() { +return this.lookupStrategy; +} + +public static SimpleAdminApiFuture newFuture( +Collection topicPartitions +) { +return AdminApiFuture.forKeys(new HashSet<>(topicPartitions)); +} + +@Override +public DeleteRecordsRequest.Builder buildBatchedRequest(int brokerId, Set keys) { +Map deletionsForTopic = new HashMap<>(); +for (Map.Entry entry: recordsToDelete.entrySet()) { +TopicPartition topicPartition = entry.getKey(); +DeleteRecordsRequestData.DeleteRecordsTopic deleteRecords = deletionsForTopic.computeIfAbsent( +topicPartition.topic(), +key -> new DeleteRecordsRequestData.DeleteRecordsTopic().setName(topicPartition.topic()) +); +deleteRecords.partitions().add(new DeleteRecordsRequestData.DeleteRecordsPartition() +.setPartitionIndex(topicPartition.partition()) +.setOffset(entry.getValue().beforeOffset())); +} + +DeleteRecordsRequestData data = new DeleteRecordsRequestData() +.setTopics(new ArrayList<>(deletionsForTopic.values())); Review Comment: Should we set timeout here like before? -- 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:
[GitHub] [kafka] showuon commented on a diff in pull request #13760: KAFKA-8982: Add retry of fetching metadata to Admin.deleteRecords
showuon commented on code in PR #13760: URL: https://github.com/apache/kafka/pull/13760#discussion_r1222617381 ## clients/src/test/java/org/apache/kafka/clients/admin/KafkaAdminClientTest.java: ## @@ -2291,6 +2289,8 @@ public void testDeleteRecords() throws Exception { try (AdminClientUnitTestEnv env = new AdminClientUnitTestEnv(cluster)) { env.kafkaClient().setNodeApiVersions(NodeApiVersions.create()); +env.kafkaClient().prepareResponse(prepareMetadataResponse(cluster, Errors.NONE)); Review Comment: Before looking into the implementation, I don't think this test is testing what we expected. The original comment is commenting around this lines: https://github.com/apache/kafka/pull/7296/files#diff-5422d10d9a7f4776c6538ae3aea27f24e94cf4ecf5e752040125aca6edc795d3R3671-R3682 And it said, when metadata response (mr) contains error, we just fail the future without retry. And here, the existing metadataResponse also fail without retry: https://github.com/apache/kafka/blob/trunk/clients/src/main/java/org/apache/kafka/clients/admin/KafkaAdminClient.java#L2969-L2972 The point is, we want to retry `metadataResponse`, not `deleteRecordResponse`. What I expected is tests like this: https://github.com/apache/kafka/blob/513e1c641d63c5e15144f9fcdafa1b56c5e5ba09/clients/src/test/java/org/apache/kafka/clients/admin/KafkaAdminClientTest.java#L5431C11-L5433 Please update the test and make sure it passed. Thanks. -- 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
[GitHub] [kafka] showuon commented on a diff in pull request #13760: KAFKA-8982: Add retry of fetching metadata to Admin.deleteRecords
showuon commented on code in PR #13760: URL: https://github.com/apache/kafka/pull/13760#discussion_r1222617381 ## clients/src/test/java/org/apache/kafka/clients/admin/KafkaAdminClientTest.java: ## @@ -2291,6 +2289,8 @@ public void testDeleteRecords() throws Exception { try (AdminClientUnitTestEnv env = new AdminClientUnitTestEnv(cluster)) { env.kafkaClient().setNodeApiVersions(NodeApiVersions.create()); +env.kafkaClient().prepareResponse(prepareMetadataResponse(cluster, Errors.NONE)); Review Comment: Before looking into the implementation, I don't think this test is testing what we expected. The original comment is commenting around this lines: https://github.com/apache/kafka/pull/7296/files#diff-5422d10d9a7f4776c6538ae3aea27f24e94cf4ecf5e752040125aca6edc795d3R3671-R3682 And it said, when metadata response (mr) contains error, we just fail the future without retry. The point is, we want to retry `metadataResponse`, not `deleteRecordResponse`. What I expected is tests like this: https://github.com/apache/kafka/blob/513e1c641d63c5e15144f9fcdafa1b56c5e5ba09/clients/src/test/java/org/apache/kafka/clients/admin/KafkaAdminClientTest.java#L5431C11-L5433 Please update the test and make sure it passed. Thanks. -- 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