aryangupta1998 commented on code in PR #8254: URL: https://github.com/apache/ozone/pull/8254#discussion_r2040373603
########## hadoop-ozone/tools/src/main/java/org/apache/hadoop/ozone/containerlog/parser/ContainerLogFileParser.java: ########## @@ -0,0 +1,222 @@ +/* + * 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.ozone.containerlog.parser; + +import java.io.BufferedReader; +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.sql.SQLException; +import java.util.ArrayList; +import java.util.List; +import java.util.concurrent.CountDownLatch; +import java.util.concurrent.ExecutorService; +import java.util.concurrent.Executors; +import java.util.concurrent.atomic.AtomicBoolean; +import java.util.stream.Collectors; +import java.util.stream.Stream; + +/** + * Parses container log files and stores container details into a database. + * Uses multithreading to process multiple log files concurrently. + */ + +public class ContainerLogFileParser { + + private ExecutorService executorService; + private static final int MAX_OBJ_IN_LIST = 5000; + + private static final String LOG_FILE_MARKER = ".log."; + private static final String LOG_LINE_SPLIT_REGEX = " \\| "; + private static final String KEY_VALUE_SPLIT_REGEX = "="; + private static final String KEY_ID = "ID"; + private static final String KEY_BCSID = "BCSID"; + private static final String KEY_STATE = "State"; + private static final String KEY_INDEX = "Index"; + private final AtomicBoolean hasErrorOccurred = new AtomicBoolean(false); + + /** + * Scans the specified log directory, processes each file in a separate thread. + * Expects each log filename to follow the format: dn-container-<roll over number>.log.<datanodeId> + * + * @param logDirectoryPath Path to the directory containing container log files. + * @param dbstore Database object used to persist parsed container data. + * @param threadCount Number of threads to use for parallel processing. + */ + + public void processLogEntries(String logDirectoryPath, ContainerDatanodeDatabase dbstore, int threadCount) + throws SQLException { + try (Stream<Path> paths = Files.walk(Paths.get(logDirectoryPath))) { + + List<Path> files = paths.filter(Files::isRegularFile).collect(Collectors.toList()); + + executorService = Executors.newFixedThreadPool(threadCount); + + CountDownLatch latch = new CountDownLatch(files.size()); + for (Path file : files) { + Path fileNamePath = file.getFileName(); + String fileName = (fileNamePath != null) ? fileNamePath.toString() : ""; + + int pos = fileName.indexOf(LOG_FILE_MARKER); + if (pos == -1) { + System.out.println("Filename format is incorrect (missing .log.): " + fileName); + continue; + } + + String datanodeId = fileName.substring(pos + 5); + + if (datanodeId.isEmpty()) { + System.out.println("Filename format is incorrect, datanodeId is missing or empty: " + fileName); + continue; + } + + executorService.submit(() -> { + + String threadName = Thread.currentThread().getName(); + try { + System.out.println(threadName + " is starting to process file: " + file.toString()); + processFile(file.toString(), dbstore, datanodeId); + } catch (Exception e) { + System.err.println("Thread " + threadName + " is stopping to process the file: " + file.toString() + + " due to SQLException: " + e.getMessage()); + hasErrorOccurred.set(true); + executorService.shutdown(); + } finally { + try { + latch.countDown(); + System.out.println(threadName + " finished processing file: " + file.toString() + + ", Latch count after countdown: " + latch.getCount()); + } catch (Exception e) { + e.printStackTrace(); + } + } + }); + } + latch.await(); + + executorService.shutdown(); + + if (hasErrorOccurred.get()) { + throw new SQLException("Log file processing failed."); + } + + } catch (IOException | InterruptedException e) { + e.printStackTrace(); + } catch (NumberFormatException e) { + System.err.println("Invalid datanode ID"); + } + } + + /** + * Processes a single container log file and extracts container details. + * Parses, batches, and writes valid container log entries into the database. + * + * @param logFilePath Path to the log file. + * @param dbstore Database object used to persist parsed container data. + * @param datanodeId Datanode ID derived from the log filename. + */ + + private void processFile(String logFilePath, ContainerDatanodeDatabase dbstore, String datanodeId) + throws SQLException { + List<DatanodeContainerInfo> batchList = new ArrayList<>(5100); Review Comment: ```suggestion List<DatanodeContainerInfo> batchList = new ArrayList<>(MAX_OBJ_IN_LIST + 100); ``` -- 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]
