arjunashok commented on code in PR #231: URL: https://github.com/apache/cassandra-sidecar/pull/231#discussion_r2246560203
########## server/src/main/java/org/apache/cassandra/sidecar/job/RepairJob.java: ########## @@ -0,0 +1,249 @@ +/* + * 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 io.vertx.core.Vertx; +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.config.RepairConfiguration; +import org.apache.cassandra.sidecar.handlers.data.RepairRequestParam; + +import static java.util.Objects.requireNonNull; + +/** + * 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 Vertx vertx; + private final RepairConfiguration repairConfiguration; + protected StorageOperations storageOperations; + + /** + * Enum representing the status of a parent repair session + */ + public enum ParentRepairStatus + { + IN_PROGRESS, COMPLETED, FAILED + } + + /** + * Constructs a job with a unique UUID, in Pending state + * + * @param vertx + * @param repairConfiguration + * @param jobId UUID representing the Job to be created + * @param storageOps + * @param repairParams + */ + public RepairJob(Vertx vertx, RepairConfiguration repairConfiguration, UUID jobId, StorageOperations storageOps, RepairRequestParam repairParams) + { + super(jobId); + this.vertx = vertx; + this.repairConfiguration = repairConfiguration; + this.storageOperations = storageOps; + this.repairParams = repairParams; + } + + @Override + public boolean isRunningOnCassandra() + { + // TODO: Leverage repair vtables to fail-fast on conflicting repairs (overlapping token-ranges or replica-sets) + // Currently does not check for concurrent repairs + return false; + } + + @Override + protected 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(), repairConfiguration.maxRepairJobRuntimeMillis()); + + int cmd = storageOperations.repair(keyspace, options); + if (cmd <= 0) + { + // repairAsync can only return 0 for replication factor 1. + LOGGER.info("Replication factor is 1. No repair is needed for keyspace '{}'", keyspace); + } + else + { + // complete the max wait time promise either when exceeding the wait time, or the result is available + Promise<Boolean> maxWaitTimePromise = Promise.promise(); + vertx.setTimer(repairConfiguration.maxRepairJobRuntimeMillis(), d -> { + LOGGER.info("Timer Poll"); + maxWaitTimePromise.tryComplete(true); + }); + + // Promise for completion of repair operation + Promise<Void> promise = Promise.promise(); + Future<Void> resultFut = promise.future(); + + // main event loop checks periodically (10s) for completion + vertx.setPeriodic(repairConfiguration.repairPollIntervalMillis(), id -> queryForCompletedRepair(promise, cmd)); + resultFut.onComplete(res -> maxWaitTimePromise.tryComplete(false)); + Future<Boolean> maxWaitTimeFut = maxWaitTimePromise.future(); + + Future<Void> compositeFut = Future.any(maxWaitTimeFut, resultFut) + // If this lambda below is evaluated, either one of the futures have completed; + // In either case, the future corresponding to the job execution is returned + .compose(f -> { + LOGGER.info("One of the futures ended waitStatus={} resultStatus={}", + maxWaitTimeFut.isComplete(), resultFut.isComplete()); + boolean isTimeout = (maxWaitTimeFut.succeeded()) ? maxWaitTimeFut.result() : false; + if (isTimeout) + { + LOGGER.error("Timer ran out before the repair job completed. Repair took too long"); + // TODO: Cancel repair? (Nice to have) + // We free up the thread (job fails) and stop polling for completion + return Future.failedFuture("Repair job taking too long"); + } + // otherwise, the result of the job is available + return resultFut; + }); + try + { + compositeFut.toCompletionStage().toCompletableFuture().get(); // wait + } + catch (Exception e) + { + throw new RuntimeException(e); + } + } + } + + /** + * {@inheritDoc} + */ + @Override + public String name() + { + return OPERATION; + } + + private Map<String, String> generateRepairOptions(RepairPayload repairPayload) + { + Map<String, String> options = new HashMap<>(); + + List<String> tables = repairPayload.tables(); + if (tables != null && !tables.isEmpty()) + { + options.put(RepairOptions.COLUMNFAMILIES.getValue(), String.join(",", tables)); + } + + Boolean isPrimaryRange = repairPayload.isPrimaryRange(); + if (isPrimaryRange != null) + { + options.put(RepairOptions.PRIMARY_RANGE.getValue(), String.valueOf(isPrimaryRange)); + } + // TODO: Verify use-cases involving multiple DCs + + String dc = repairPayload.datacenter(); + if (dc != null) + { + options.put(RepairOptions.DATACENTERS.getValue(), dc); + } + + List<String> hosts = repairPayload.hosts(); + if (hosts != null && !hosts.isEmpty()) + { + options.put(RepairOptions.HOSTS.getValue(), String.join(",", requireNonNull(hosts))); + } + + if (repairPayload.startToken() != null && repairPayload.endToken() != null) + { + options.put(RepairOptions.RANGES.getValue(), repairPayload.startToken() + ":" + repairPayload.endToken()); + } + + if (repairPayload.repairType() == RepairPayload.RepairType.INCREMENTAL) + { + options.put(RepairOptions.INCREMENTAL.getValue(), Boolean.TRUE.toString()); + } + + if (repairPayload.force() != null) + { + options.put(RepairOptions.FORCE_REPAIR.getValue(), String.valueOf(repairPayload.force())); + } + + if (repairPayload.isValidate() != null) + { + options.put(RepairOptions.PREVIEW.getValue(), PREVIEW_KIND_REPAIRED); + } + return options; + } + + private void queryForCompletedRepair(Promise<Void> promise, int cmd) + { + LOGGER.info("Polling repair operation for jobId={} status={}", this.jobId(), promise.future().isComplete()); + List<String> status = storageOperations.getParentRepairStatus(cmd); + String queriedString = "queried for parent session status and"; + if (status == null) + { + LOGGER.error("{} couldn't find repair status for cmd: {}", Review Comment: Good point. Makes sense to fail in this case. ########## server/src/main/java/org/apache/cassandra/sidecar/job/RepairJob.java: ########## @@ -0,0 +1,249 @@ +/* + * 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 io.vertx.core.Vertx; +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.config.RepairConfiguration; +import org.apache.cassandra.sidecar.handlers.data.RepairRequestParam; + +import static java.util.Objects.requireNonNull; + +/** + * 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 Vertx vertx; + private final RepairConfiguration repairConfiguration; + protected StorageOperations storageOperations; + + /** + * Enum representing the status of a parent repair session + */ + public enum ParentRepairStatus + { + IN_PROGRESS, COMPLETED, FAILED + } + + /** + * Constructs a job with a unique UUID, in Pending state + * + * @param vertx + * @param repairConfiguration + * @param jobId UUID representing the Job to be created + * @param storageOps + * @param repairParams + */ + public RepairJob(Vertx vertx, RepairConfiguration repairConfiguration, UUID jobId, StorageOperations storageOps, RepairRequestParam repairParams) + { + super(jobId); + this.vertx = vertx; + this.repairConfiguration = repairConfiguration; + this.storageOperations = storageOps; + this.repairParams = repairParams; + } + + @Override + public boolean isRunningOnCassandra() + { + // TODO: Leverage repair vtables to fail-fast on conflicting repairs (overlapping token-ranges or replica-sets) + // Currently does not check for concurrent repairs + return false; + } + + @Override + protected 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(), repairConfiguration.maxRepairJobRuntimeMillis()); + + int cmd = storageOperations.repair(keyspace, options); + if (cmd <= 0) + { + // repairAsync can only return 0 for replication factor 1. + LOGGER.info("Replication factor is 1. No repair is needed for keyspace '{}'", keyspace); + } + else + { + // complete the max wait time promise either when exceeding the wait time, or the result is available + Promise<Boolean> maxWaitTimePromise = Promise.promise(); + vertx.setTimer(repairConfiguration.maxRepairJobRuntimeMillis(), d -> { + LOGGER.info("Timer Poll"); + maxWaitTimePromise.tryComplete(true); + }); + + // Promise for completion of repair operation + Promise<Void> promise = Promise.promise(); + Future<Void> resultFut = promise.future(); + + // main event loop checks periodically (10s) for completion + vertx.setPeriodic(repairConfiguration.repairPollIntervalMillis(), id -> queryForCompletedRepair(promise, cmd)); + resultFut.onComplete(res -> maxWaitTimePromise.tryComplete(false)); + Future<Boolean> maxWaitTimeFut = maxWaitTimePromise.future(); + + Future<Void> compositeFut = Future.any(maxWaitTimeFut, resultFut) + // If this lambda below is evaluated, either one of the futures have completed; + // In either case, the future corresponding to the job execution is returned + .compose(f -> { + LOGGER.info("One of the futures ended waitStatus={} resultStatus={}", + maxWaitTimeFut.isComplete(), resultFut.isComplete()); + boolean isTimeout = (maxWaitTimeFut.succeeded()) ? maxWaitTimeFut.result() : false; + if (isTimeout) + { + LOGGER.error("Timer ran out before the repair job completed. Repair took too long"); + // TODO: Cancel repair? (Nice to have) + // We free up the thread (job fails) and stop polling for completion + return Future.failedFuture("Repair job taking too long"); + } + // otherwise, the result of the job is available + return resultFut; + }); + try + { + compositeFut.toCompletionStage().toCompletableFuture().get(); // wait Review Comment: With the changes made addressing Yifan's comment that links the PeriodicTask's promise to be handled when the internal future is complete (returned from `executeInternal`), it should address this concern. Let me know if that is what you are referring to. ########## server/src/main/java/org/apache/cassandra/sidecar/job/RepairJob.java: ########## @@ -0,0 +1,249 @@ +/* + * 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 io.vertx.core.Vertx; +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.config.RepairConfiguration; +import org.apache.cassandra.sidecar.handlers.data.RepairRequestParam; + +import static java.util.Objects.requireNonNull; + +/** + * 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 Vertx vertx; + private final RepairConfiguration repairConfiguration; + protected StorageOperations storageOperations; + + /** + * Enum representing the status of a parent repair session + */ + public enum ParentRepairStatus + { + IN_PROGRESS, COMPLETED, FAILED + } + + /** + * Constructs a job with a unique UUID, in Pending state + * + * @param vertx + * @param repairConfiguration + * @param jobId UUID representing the Job to be created + * @param storageOps + * @param repairParams + */ + public RepairJob(Vertx vertx, RepairConfiguration repairConfiguration, UUID jobId, StorageOperations storageOps, RepairRequestParam repairParams) + { + super(jobId); + this.vertx = vertx; + this.repairConfiguration = repairConfiguration; + this.storageOperations = storageOps; + this.repairParams = repairParams; + } + + @Override + public boolean isRunningOnCassandra() + { + // TODO: Leverage repair vtables to fail-fast on conflicting repairs (overlapping token-ranges or replica-sets) + // Currently does not check for concurrent repairs + return false; + } + + @Override + protected 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(), repairConfiguration.maxRepairJobRuntimeMillis()); + + int cmd = storageOperations.repair(keyspace, options); + if (cmd <= 0) + { + // repairAsync can only return 0 for replication factor 1. + LOGGER.info("Replication factor is 1. No repair is needed for keyspace '{}'", keyspace); + } + else + { + // complete the max wait time promise either when exceeding the wait time, or the result is available + Promise<Boolean> maxWaitTimePromise = Promise.promise(); + vertx.setTimer(repairConfiguration.maxRepairJobRuntimeMillis(), d -> { + LOGGER.info("Timer Poll"); + maxWaitTimePromise.tryComplete(true); + }); + + // Promise for completion of repair operation + Promise<Void> promise = Promise.promise(); + Future<Void> resultFut = promise.future(); + + // main event loop checks periodically (10s) for completion + vertx.setPeriodic(repairConfiguration.repairPollIntervalMillis(), id -> queryForCompletedRepair(promise, cmd)); + resultFut.onComplete(res -> maxWaitTimePromise.tryComplete(false)); + Future<Boolean> maxWaitTimeFut = maxWaitTimePromise.future(); + + Future<Void> compositeFut = Future.any(maxWaitTimeFut, resultFut) + // If this lambda below is evaluated, either one of the futures have completed; + // In either case, the future corresponding to the job execution is returned + .compose(f -> { + LOGGER.info("One of the futures ended waitStatus={} resultStatus={}", + maxWaitTimeFut.isComplete(), resultFut.isComplete()); + boolean isTimeout = (maxWaitTimeFut.succeeded()) ? maxWaitTimeFut.result() : false; + if (isTimeout) + { + LOGGER.error("Timer ran out before the repair job completed. Repair took too long"); + // TODO: Cancel repair? (Nice to have) + // We free up the thread (job fails) and stop polling for completion + return Future.failedFuture("Repair job taking too long"); Review Comment: So, the `asyncResult` implementation is meant for tracking the execution of a OperationalJob and it's timeout from the handler's standpoint, while `executeInternal` is specific to the actual job's implementation, where we want to wait based on the job-specific configuration for the asynchronous command. Also, the timeouts are handled differently - the main job handler succeeds on a job timeout, but the repair job's async repair command fails on the repair timeout. There are similarities in the implementation, but I figured it is simpler to allow jobs to manage their own timeouts independently. -- 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
