nizhikov commented on a change in pull request #8909: URL: https://github.com/apache/ignite/pull/8909#discussion_r602062688
########## File path: modules/core/src/main/java/org/apache/ignite/internal/cdc/IgniteCDC.java ########## @@ -0,0 +1,485 @@ +/* + * 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.ignite.internal.cdc; + +import java.io.File; +import java.io.IOException; +import java.io.Serializable; +import java.nio.file.Files; +import java.nio.file.Path; +import java.nio.file.Paths; +import java.util.ArrayList; +import java.util.Comparator; +import java.util.HashSet; +import java.util.Iterator; +import java.util.List; +import java.util.Set; +import java.util.UUID; +import java.util.function.Predicate; +import java.util.stream.Stream; +import org.apache.ignite.IgniteCheckedException; +import org.apache.ignite.IgniteException; +import org.apache.ignite.IgniteLogger; +import org.apache.ignite.cdc.CaptureDataChangeConfiguration; +import org.apache.ignite.cdc.CaptureDataChangeConsumer; +import org.apache.ignite.cdc.ChangeEvent; +import org.apache.ignite.configuration.DataStorageConfiguration; +import org.apache.ignite.configuration.IgniteConfiguration; +import org.apache.ignite.internal.IgnitionEx; +import org.apache.ignite.internal.MarshallerContextImpl; +import org.apache.ignite.internal.pagemem.wal.WALIterator; +import org.apache.ignite.internal.processors.cache.binary.CacheObjectBinaryProcessorImpl; +import org.apache.ignite.internal.processors.cache.persistence.filename.PdsConsistentIdProcessor; +import org.apache.ignite.internal.processors.cache.persistence.wal.WALPointer; +import org.apache.ignite.internal.processors.cache.persistence.wal.reader.IgniteWalIteratorFactory; +import org.apache.ignite.internal.util.typedef.F; +import org.apache.ignite.internal.util.typedef.internal.CU; +import org.apache.ignite.internal.util.typedef.internal.U; +import org.apache.ignite.lang.IgniteBiTuple; + +import static org.apache.ignite.configuration.DataStorageConfiguration.DFLT_CDC_PATH; +import static org.apache.ignite.internal.pagemem.wal.record.WALRecord.RecordType.DATA_RECORD_V2; +import static org.apache.ignite.internal.processors.cache.persistence.filename.PdsConsistentIdProcessor.NODE_PATTERN; +import static org.apache.ignite.internal.processors.cache.persistence.filename.PdsConsistentIdProcessor.UUID_STR_PATTERN; +import static org.apache.ignite.internal.processors.cache.persistence.wal.FileWriteAheadLogManager.WAL_NAME_PATTERN; + +/** + * CDC(Capture Data Change) application. + * Application run independently of Ignite node process and provide ability for the {@link CaptureDataChangeConsumer} to consume events({@link ChangeEvent}) from WAL segments. + * User should responsible {@link CaptureDataChangeConsumer} implementation with custom consumption logic. + * + * Ignite node should be explicitly configured for using {@link IgniteCDC}. + * <ol> + * <li>Set {@link DataStorageConfiguration#setCdcEnabled(boolean)} to true.</li> + * <li>Optional: Set {@link DataStorageConfiguration#setCdcPath(String)} to path to the directory to store WAL setgments for CDC.</li> + * <li>Optional: Set {@link DataStorageConfiguration#setWalForceArchiveTimeout(long)} to configure timeout for force WAL rollover, + * so new events will be available for consumptions with the predicted time.</li> + * </ol> + * + * When {@link DataStorageConfiguration#getCdcPath()} is true then Ignite node on each WAL segment rollover creates hard link + * to archive WAL segment in {@link DataStorageConfiguration#getCdcPath()} directory. + * {@link IgniteCDC} application takes segment file and consumes events from it. After successfull consumption (see {@link CaptureDataChangeConsumer#onChange(Iterator)}) + * WAL segement will be deleted from directory. + * + * Several Ignite nodes can be started on the same host. + * If your deployment done with custom consistent id then you should specify it via {@link IgniteConfiguration#setConsistentId(Serializable)} in provided {@link IgniteConfiguration}. + * Otherwise you should specify auto generated consistent id and node index with the following options: + * <ul> + * <li>{@link CaptureDataChangeConfiguration#setAutoGeneratedConsistentId(UUID)}.</li> + * <li>{@link CaptureDataChangeConfiguration#setNodeIndex(int)}.</li> + * </ul> + * + * Application works as follows: + * <ol> + * <li>Search node work directory based on provided {@link IgniteConfiguration} and system properties.</li> + * <li>Await for creation of CDC directory if it not exists.</li> + * <li>Acquire file lock to ensure exclusive consumption.</li> + * <li>Loads state of consumption if it exists.</li> + * <li>Infinetely wait for new available segement and process it.</li> + * </ol> + * + * @see DataStorageConfiguration#setCdcEnabled(boolean) + * @see DataStorageConfiguration#setCdcPath(String) + * @see DataStorageConfiguration#setWalForceArchiveTimeout(long) + * @see CommandLineStartup + * @see CaptureDataChangeConsumer + * @see DataStorageConfiguration#DFLT_CDC_PATH + */ +public class IgniteCDC implements Runnable { + /** State dir. */ + public static final String STATE_DIR = "state"; + + /** Ignite configuration. */ + private final IgniteConfiguration cfg; + + /** CDC configuration. */ + private final CaptureDataChangeConfiguration cdcCfg; + + /** WAL iterator factory. */ + private final IgniteWalIteratorFactory factory; + + /** Events consumer. */ + private final WALRecordsConsumer<?, ?> consumer; + + /** Logger. */ + private final IgniteLogger log; + + /** CDC directory. */ + private Path cdcDir; + + /** Binary meta directory. */ + private File binaryMeta; + + /** Marshaller directory. */ + private File marshaller; + + /** CDC state. */ + private CDCConsumerState state; + + /** Save state to start from. */ + private WALPointer initState; + + /** Consistent ID. */ + private final String nodeDir; + + /** Previous segments. */ + private final List<Path> prevSegments = new ArrayList<>(); + + /** + * @param cfg Ignite configuration. + * @param cdcCfg CDC configuration. + */ + public IgniteCDC(IgniteConfiguration cfg, CaptureDataChangeConfiguration cdcCfg) { + this.cfg = cfg; + this.cdcCfg = cdcCfg; + this.consumer = new WALRecordsConsumer<>(cdcCfg.getConsumer()); + + log = logger(cfg, workDir(cfg)); + factory = new IgniteWalIteratorFactory(log); + + if (!CU.isPersistenceEnabled(cfg)) { + log.error("Persistence disabled. IgniteCDC can't run!"); + + throw new IllegalArgumentException("Persistence disabled. IgniteCDC can't run!"); + } + + nodeDir = consistentId(cfg); + + if (nodeDir == null) { + log.warning("Can't determine nodeDir. It is recommended to set Consistent ID for production " + + "clusters(use IgniteConfiguration.setConsistentId or CaptureDataChangeConfiguration.setConsistentId " + + "and CaptureDataChangeConfiguration.setNodeIndex)."); + } + } + + /** Runs CDC. */ + @Override public void run() { + try { + runX(); + } + catch (Throwable e) { + e.printStackTrace(); + + throw new RuntimeException(e); + } + } + + /** Runs CDC application with possible exception. */ + public void runX() throws Exception { + if (log.isInfoEnabled()) { + log.info("Starting Ignite CDC Application."); + log.info("Consumer -\t" + consumer.toString()); + } + + cdcDir = findCDCDir(workDir(cfg)); + + try (CDCFileLockHolder lock = + new CDCFileLockHolder(cdcDir.toString(), () -> "cdc.lock", log)) { Review comment: > use the id of running consumer instance instead of cdc-lock Right now we don't have such a thing as "id", because only one concurrent consumer supported. :) -- 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: [email protected]
