mumrah commented on a change in pull request #9100: URL: https://github.com/apache/kafka/pull/9100#discussion_r478067221
########## File path: core/src/main/scala/kafka/server/AlterIsrManager.scala ########## @@ -0,0 +1,145 @@ +/** + * 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 kafka.server + +import java.util +import java.util.concurrent.{ScheduledFuture, TimeUnit} +import java.util.concurrent.atomic.AtomicLong + +import kafka.api.LeaderAndIsr +import kafka.metrics.KafkaMetricsGroup +import kafka.utils.{Logging, Scheduler} +import kafka.zk.KafkaZkClient +import org.apache.kafka.clients.ClientResponse +import org.apache.kafka.common.TopicPartition +import org.apache.kafka.common.message.AlterIsrRequestData.{AlterIsrRequestPartitions, AlterIsrRequestTopics} +import org.apache.kafka.common.message.{AlterIsrRequestData, AlterIsrResponseData} +import org.apache.kafka.common.protocol.Errors +import org.apache.kafka.common.requests.{AlterIsrRequest, AlterIsrResponse} +import org.apache.kafka.common.utils.Time + +import scala.collection.mutable +import scala.jdk.CollectionConverters._ + +/** + * Handles the sending of AlterIsr requests to the controller. Updating the ISR is an asynchronous operation, + * so partitions will learn about updates through LeaderAndIsr messages sent from the controller + */ +trait AlterIsrManager { + def enqueueIsrUpdate(alterIsrItem: AlterIsrItem): Unit + + def clearPending(topicPartition: TopicPartition): Unit +} + +case class AlterIsrItem(topicPartition: TopicPartition, leaderAndIsr: LeaderAndIsr, callback: Errors => Unit) + +class AlterIsrManagerImpl(val controllerChannelManager: BrokerToControllerChannelManager, + val zkClient: KafkaZkClient, + val scheduler: Scheduler, + val time: Time, + val brokerId: Int) extends AlterIsrManager with Logging with KafkaMetricsGroup { + + private val unsentIsrUpdates: mutable.Map[TopicPartition, AlterIsrItem] = new mutable.HashMap[TopicPartition, AlterIsrItem]() + private val lastIsrChangeMs = new AtomicLong(0) + private val lastIsrPropagationMs = new AtomicLong(0) + + @volatile private var scheduledRequest: Option[ScheduledFuture[_]] = None + + override def enqueueIsrUpdate(alterIsrItem: AlterIsrItem): Unit = { + unsentIsrUpdates synchronized { + unsentIsrUpdates(alterIsrItem.topicPartition) = alterIsrItem + lastIsrChangeMs.set(time.milliseconds) + // Rather than sending right away, we'll delay at most 50ms to allow for batching of ISR changes happening + // in fast succession + if (scheduledRequest.isEmpty) { Review comment: I think that sounds pretty reasonable. Would we need any kind of timeout at this layer, or just rely on the underlying channel to provide timeouts? ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
