arjunashok commented on code in PR #231: URL: https://github.com/apache/cassandra-sidecar/pull/231#discussion_r2253288265
########## server/src/main/java/org/apache/cassandra/sidecar/job/RepairJob.java: ########## @@ -0,0 +1,244 @@ +/* + * 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.cassandra.sidecar.job; + +import java.io.IOException; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.UUID; + +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import io.vertx.core.Future; +import io.vertx.core.Promise; +import org.apache.cassandra.sidecar.adapters.base.RepairOptions; +import org.apache.cassandra.sidecar.common.request.data.RepairPayload; +import org.apache.cassandra.sidecar.common.server.StorageOperations; +import org.apache.cassandra.sidecar.concurrent.TaskExecutorPool; +import org.apache.cassandra.sidecar.config.RepairJobsConfiguration; +import org.apache.cassandra.sidecar.handlers.data.RepairRequestParam; +import org.jetbrains.annotations.NotNull; + +/** + * Implementation of {@link OperationalJob} to perform repair operation. + */ +public class RepairJob extends OperationalJob +{ + private static final Logger LOGGER = LoggerFactory.getLogger(RepairJob.class); + private static final String OPERATION = "repair"; + private static final String PREVIEW_KIND_REPAIRED = "REPAIRED"; + + private final RepairRequestParam repairParams; + private final RepairJobsConfiguration config; + private final TaskExecutorPool internalPool; + protected StorageOperations storageOperations; + + /** + * Enum representing the status of a parent repair session + */ + public enum ParentRepairStatus + { + IN_PROGRESS, COMPLETED, FAILED + } + + /** + * Constructor for creation of RepairJob + * + * @param taskExecutorPool TaskExecutorPool instance (for testing) + * @param config Repair job configuration + * @param jobId UUID representing the Job to be created + * @param storageOps Reference to the storage operations interface + * @param repairParams Repair request parameters + */ + public RepairJob(TaskExecutorPool taskExecutorPool, + RepairJobsConfiguration config, + UUID jobId, + StorageOperations storageOps, + RepairRequestParam repairParams) + { + super(jobId); + this.internalPool = taskExecutorPool; + this.config = config; + this.storageOperations = storageOps; + this.repairParams = repairParams; + } + + /** + * {@inheritDoc} + * RepairJob allows parallel executions since multiple repairs can run concurrently + * on different tables or with different parameters. + */ + @Override + public boolean hasConflict(@NotNull List<OperationalJob> sameOperationJobs) + { + // For now, we simply allow all repair jobs to run in parallel + // A more sophisticated implementation could check the repair parameters + // against currently running repairs to detect potential conflicts + return false; + } + + @Override + protected Future<Void> executeInternal() + { + Map<String, String> options = generateRepairOptions(repairParams.requestpayload()); + String keyspace = repairParams.keyspace().name(); + + LOGGER.info("Executing repair operation for keyspace {} jobId={} maxRuntime={}", + keyspace, this.jobId(), config.maxRepairJobRuntime()); + + int cmd = storageOperations.repair(keyspace, options); + if (cmd <= 0) + { + // When there are no relevant token ranges for the keyspace or RF is 1, the repair is inapplicable + LOGGER.info("Repair is not applicable for the provided options and keyspace '{}' jobId '{}'", keyspace, this.jobId()); + return Future.succeededFuture(); + } + Promise<Void> repairPromise = Promise.promise(); + queryForCompletedRepair(repairPromise, cmd); + + if (!repairPromise.future().isComplete()) + { + Promise<Boolean> maxWaitTimePromise = Promise.promise(); + long timerId = internalPool.setTimer(config.maxRepairJobRuntime().toMillis(), + d -> maxWaitTimePromise.tryComplete(true)); + + long periodicId = internalPool.setPeriodic(config.repairPollInterval().toMillis(), + id -> queryForCompletedRepair(repairPromise, cmd)); + + repairPromise.future().onComplete(ar -> { + internalPool.cancelTimer(timerId); + internalPool.cancelTimer(periodicId); + maxWaitTimePromise.tryComplete(false); + }); + + return Future.any(maxWaitTimePromise.future(), repairPromise.future()) + .compose(f -> { + boolean isTimeout = maxWaitTimePromise.future().result(); + if (isTimeout) + { + LOGGER.error("Timer ran out before the repair job completed. Repair took too long"); + return Future.failedFuture("Repair job taking too long"); Review Comment: Yep, valid concern. The current implementation uses a timeout to prevent the sidecar from holding thread resources indefinitely but simply failing the future doesn't provide a good way for clients to continue tracking the repair operation that's still running in Cassandra. How about a hybrid approach - Keep the timeout mechanism to release sidecar resources after maxRepairJobRuntime, but change how we handle the timeout. Instead of marking the job as failed on timeout, transition it to a 'DETACHED' state that indicates 'no longer actively monitored by sidecar, but still running in Cassandra'. Additionally, if we have access to Cassandra's repair session/command ID, it could added to the job response, allowing clients to query/troubleshoot with Cassandra directly for status if needed. -- 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: pr-unsubscr...@cassandra.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: pr-unsubscr...@cassandra.apache.org For additional commands, e-mail: pr-h...@cassandra.apache.org
