sarvekshayr commented on code in PR #10673: URL: https://github.com/apache/ozone/pull/10673#discussion_r3549645420
########## hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/container/export/ContainerExportManager.java: ########## @@ -0,0 +1,499 @@ +/* + * 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.hadoop.hdds.scm.container.export; + +import java.io.BufferedWriter; +import java.io.File; +import java.io.IOException; +import java.nio.charset.StandardCharsets; +import java.nio.file.Files; +import java.nio.file.Path; +import java.nio.file.Paths; +import java.time.Instant; +import java.time.ZoneOffset; +import java.time.format.DateTimeFormatter; +import java.util.ArrayList; +import java.util.List; +import java.util.Map; +import java.util.UUID; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.ExecutorService; +import java.util.concurrent.Executors; +import java.util.concurrent.Future; +import java.util.concurrent.TimeUnit; +import org.apache.commons.io.FileUtils; +import org.apache.hadoop.hdds.conf.OzoneConfiguration; +import org.apache.hadoop.hdds.protocol.proto.HddsProtos.LifeCycleState; +import org.apache.hadoop.hdds.scm.container.ContainerHealthState; +import org.apache.hadoop.hdds.scm.container.ContainerID; +import org.apache.hadoop.hdds.scm.container.ContainerManager; +import org.apache.hadoop.hdds.server.ServerUtils; +import org.apache.hadoop.hdds.utils.Archiver; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +/** + * Manages asynchronous container ID export jobs on SCM leader. + */ +public class ContainerExportManager { + + private static final Logger LOG = LoggerFactory.getLogger(ContainerExportManager.class); + + private static final DateTimeFormatter METADATA_TIMESTAMP_FORMAT = + DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm:ss'Z'").withZone(ZoneOffset.UTC); + private static final DateTimeFormatter FILENAME_TIMESTAMP_FORMAT = + DateTimeFormatter.ofPattern("yyyyMMdd'T'HHmmss'Z'").withZone(ZoneOffset.UTC); + + private static final String EXPORT_SUBDIR = "exports"; + private static final int DEFAULT_SHARD_SIZE = 500_000; + private static final int DEFAULT_PAGE_SIZE = 100_000; + private static final int MAX_SHARD_SIZE = 5_000_000; + private static final int MAX_PAGE_SIZE = 1_000_000; + + private final Map<String, ExportJob> jobTracker = new ConcurrentHashMap<>(); + private final ExecutorService workerPool; + private final Map<String, Future<?>> runningTasks = new ConcurrentHashMap<>(); + private final ContainerManager containerManager; + private final String exportDirectory; + private final int defaultShardSize; + private final int defaultPageSize; + private final Object submitLock = new Object(); + + public ContainerExportManager(ContainerManager containerManager, + OzoneConfiguration conf) { + this(containerManager, resolveExportDirectory(conf), + DEFAULT_SHARD_SIZE, DEFAULT_PAGE_SIZE); + } + + ContainerExportManager(ContainerManager containerManager, + String exportDirectory, int defaultShardSize, int defaultPageSize) { + this.containerManager = containerManager; + this.exportDirectory = exportDirectory; + this.defaultShardSize = defaultShardSize; + this.defaultPageSize = defaultPageSize; + this.workerPool = Executors.newSingleThreadExecutor(r -> { + Thread t = new Thread(r, "ContainerExportWorker"); + t.setDaemon(true); + return t; + }); + + try { + Files.createDirectories(Paths.get(exportDirectory)); + } catch (IOException e) { + LOG.error("Failed to create export directory: {}", exportDirectory, e); + } + LOG.info("ContainerExportManager initialized (dir={}, defaultShardSize={}, defaultPageSize={})", + exportDirectory, defaultShardSize, defaultPageSize); + } + + private static String resolveExportDirectory(OzoneConfiguration conf) { + File scmDbDir = ServerUtils.getScmDbDir(conf); + return new File(scmDbDir, EXPORT_SUBDIR).getAbsolutePath(); + } + + /** + * Submit a container ID export job. + * + * @param start optional inclusive start container ID (0 for beginning) + * @param lifeCycleState optional lifecycle filter + * @param healthState optional health filter + * @param maxRows optional row limit (0 = unlimited) + * @param pageSize IDs fetched per SCM read (0 = manager default) + * @param shardSize IDs per TAR entry (0 = manager default) + * @return job id + */ + public String submitJob(ContainerID start, LifeCycleState lifeCycleState, + ContainerHealthState healthState, long maxRows, int pageSize, int shardSize) { + if (lifeCycleState == null && healthState == null) { + throw new IllegalArgumentException("At least one of healthState or lifecycleState filter is required."); + } + validateRequest(start, maxRows, pageSize, shardSize); + int resolvedPageSize = pageSize > 0 ? pageSize : defaultPageSize; + int resolvedShardSize = shardSize > 0 ? shardSize : defaultShardSize; + + String jobId = UUID.randomUUID().toString(); + String scope = buildScope(lifeCycleState, healthState); + Instant now = Instant.now(); + String metadataTimestamp = METADATA_TIMESTAMP_FORMAT.format(now); + String fileTimestamp = FILENAME_TIMESTAMP_FORMAT.format(now); + String tarFileName = String.format("container-ids-%s-%s-%s.tar", scope, fileTimestamp, jobId); + String tarPath = exportDirectory + File.separator + tarFileName; + + ExportJob job = new ExportJob(jobId, scope, metadataTimestamp, tarPath, + start, lifeCycleState, healthState, maxRows, resolvedPageSize, resolvedShardSize); + + synchronized (submitLock) { + boolean exportInProgress = jobTracker.values().stream() + .anyMatch(j -> j.getState() == ContainerExportStatus.State.RUNNING); + if (exportInProgress) { + throw new IllegalStateException("Another container ID export is already running."); + } + jobTracker.put(jobId, job); + } + + Future<?> future = workerPool.submit(() -> executeExport(job)); + runningTasks.put(jobId, future); + LOG.info("Submitted container ID export job {} (scope={}, start={}, maxRows={}, pageSize={}, shardSize={})", + jobId, scope, start, maxRows, resolvedPageSize, resolvedShardSize); + return jobId; + } + + private static void validateRequest(ContainerID start, long maxRows, int pageSize, int shardSize) { + if (start != null && start.getProtobuf().getId() < 0) { + throw new IllegalArgumentException("start container ID must be non-negative."); + } + if (maxRows < 0) { + throw new IllegalArgumentException("maxRows must be non-negative."); + } + if (pageSize < 0 || pageSize > MAX_PAGE_SIZE) { + throw new IllegalArgumentException("pageSize must be between 0 and " + MAX_PAGE_SIZE + "."); + } + if (shardSize < 0 || shardSize > MAX_SHARD_SIZE) { + throw new IllegalArgumentException("shardSize must be between 0 and " + MAX_SHARD_SIZE + "."); + } + } + + public ContainerExportStatus getJobStatus(String jobId) { + ExportJob job = jobTracker.get(jobId); + if (job == null) { + return null; + } + return job.toStatus(); + } + + public void shutdown() { + LOG.info("Shutting down ContainerExportManager"); + workerPool.shutdownNow(); + try { + workerPool.awaitTermination(30, TimeUnit.SECONDS); + } catch (InterruptedException e) { + LOG.warn("Timeout waiting for export worker shutdown", e); + Thread.currentThread().interrupt(); + } + runningTasks.clear(); + } + + Map<String, ExportJob> getJobTracker() { + return jobTracker; + } + + private void executeExport(ExportJob job) { + Path jobDir = Paths.get(exportDirectory, job.getJobId()); + Path workDir = jobDir.resolve("work"); + File tarFile = new File(job.getTarPath()); + long startTimeMs = System.currentTimeMillis(); + job.setStartTimeMs(startTimeMs); + + try { + Files.createDirectories(workDir); + job.setState(ContainerExportStatus.State.RUNNING); + + ContainerID cursor = job.getStartContainerId(); + int pageSize = job.getPageSize(); + int shardSize = job.getShardSize(); + int fileIndex = 1; + long totalRows = 0; + long recordsInCurrentFile = 0; + BufferedWriter writer = null; + Path currentShardPath = null; + // Pre-allocated buffer: ~12 chars per ID (up to 20 digits + newline) per page. + StringBuilder buf = new StringBuilder(pageSize * 12); + + try { + while (true) { + if (Thread.currentThread().isInterrupted()) { + throw new InterruptedException("Job cancelled"); + } + + int fetchCount = pageSize; + if (job.getMaxRows() > 0) { + long remaining = job.getMaxRows() - totalRows; + if (remaining <= 0) { + break; + } + fetchCount = (int) Math.min(fetchCount, remaining); + } + + List<ContainerID> page = containerManager.getContainerIDs( + cursor, fetchCount, job.getLifeCycleState(), job.getHealthState()); Review Comment: The final TAR would either have the previous state from last RM cycle or current state of the running RM cycle, which is acceptable. Updated the same in the cli option description. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
