ChenSammi commented on code in PR #8661: URL: https://github.com/apache/ozone/pull/8661#discussion_r2185085185
########## hadoop-ozone/integration-test/src/test/java/org/apache/hadoop/ozone/scm/node/TestVolumeChoosingPolicy.java: ########## @@ -0,0 +1,252 @@ +/* + * 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.scm.node; + +import static org.junit.jupiter.api.Assertions.assertTrue; +import static org.mockito.Mockito.mock; + +import java.io.IOException; +import java.nio.file.Files; +import java.nio.file.Path; +import java.time.Duration; +import java.util.ArrayList; +import java.util.List; +import java.util.Map; +import java.util.Random; +import java.util.UUID; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.CountDownLatch; +import java.util.concurrent.ExecutorService; +import java.util.concurrent.Executors; +import java.util.concurrent.TimeUnit; +import java.util.concurrent.atomic.AtomicInteger; +import java.util.concurrent.atomic.AtomicLong; +import org.apache.commons.lang3.tuple.Pair; +import org.apache.hadoop.hdds.conf.OzoneConfiguration; +import org.apache.hadoop.hdds.fs.MockSpaceUsageCheckFactory; +import org.apache.hadoop.hdds.fs.MockSpaceUsageSource; +import org.apache.hadoop.hdds.fs.SpaceUsageCheckFactory; +import org.apache.hadoop.hdds.fs.SpaceUsagePersistence; +import org.apache.hadoop.hdds.fs.SpaceUsageSource; +import org.apache.hadoop.hdds.scm.ScmConfigKeys; +import org.apache.hadoop.ozone.container.common.statemachine.StateContext; +import org.apache.hadoop.ozone.container.common.volume.HddsVolume; +import org.apache.hadoop.ozone.container.common.volume.MutableVolumeSet; +import org.apache.hadoop.ozone.container.common.volume.StorageVolume; +import org.apache.hadoop.ozone.container.diskbalancer.policy.DefaultVolumeChoosingPolicy; +import org.apache.hadoop.ozone.container.diskbalancer.policy.VolumeChoosingPolicy; +import org.junit.jupiter.api.AfterEach; +import org.junit.jupiter.api.BeforeEach; +import org.junit.jupiter.api.Test; +import org.junit.jupiter.api.Timeout; +import org.junit.jupiter.api.io.TempDir; + +/** + * This class tests the VolumeChoosingPolicy. + */ +public class TestVolumeChoosingPolicy { + + private static final int NUM_VOLUMES = 20; + private static final int NUM_THREADS = 10; + private static final int NUM_ITERATIONS = 10000; + private static final double THRESHOLD = 0.1; // 10% threshold + + private static final OzoneConfiguration CONF = new OzoneConfiguration(); + + @TempDir + private Path baseDir; + + private MutableVolumeSet volumeSet; + private List<HddsVolume> hddsVolumes; + private VolumeChoosingPolicy volumeChoosingPolicy; + private ExecutorService executor; + + // delta sizes for source volumes + private Map<HddsVolume, Long> deltaSizes = new ConcurrentHashMap<>(); + + @BeforeEach + public void setup() throws Exception { + hddsVolumes = new ArrayList<>(); + initialiseVolumeSet(); + createVolumes(); + volumeChoosingPolicy = new DefaultVolumeChoosingPolicy(); + executor = Executors.newFixedThreadPool(NUM_THREADS); + } + + @AfterEach + public void cleanUp() { + hddsVolumes.forEach(HddsVolume::shutdown); + + // Shutdown executor service + if (executor != null && !executor.isShutdown()) { + executor.shutdownNow(); + } + + // Clear in-progress delta sizes + if (deltaSizes != null) { + deltaSizes.clear(); + } + + // Clear VolumeSet + if (volumeSet != null) { + volumeSet.shutdown(); + volumeSet = null; + } + } + + @Test + @Timeout(300) + public void testConcurrentVolumeChoosingPerformance() throws Exception { + testPolicyPerformance("DefaultVolumeChoosingPolicy", volumeChoosingPolicy); + } + + /** + * pairChosenCount: Number of successful volume pair choices from the policy. + * FailureCount: Failures due to any exceptions thrown during volume choice or null return. + */ + private void testPolicyPerformance(String policyName, VolumeChoosingPolicy policy) throws Exception { + CountDownLatch latch = new CountDownLatch(NUM_THREADS); + AtomicInteger pairChosenCount = new AtomicInteger(0); + AtomicInteger pairNotChosenCount = new AtomicInteger(0); + AtomicInteger failureCount = new AtomicInteger(0); + AtomicLong totalTimeNanos = new AtomicLong(0); + + Random rand = new Random(); + + for (int i = 0; i < NUM_THREADS; i++) { + executor.submit(() -> { + try { + long threadStart = System.nanoTime(); + int volumeChosen = 0; + int volumeNotChosen = 0; + int failures = 0; + + for (int j = 0; j < NUM_ITERATIONS; j++) { + if (rand.nextDouble() < 0.05 && hddsVolumes.size() >= 2) { + HddsVolume sourceVolume = hddsVolumes.get(rand.nextInt(hddsVolumes.size())); + + // Randomly pick a destination volume, ensuring it's different from the source + HddsVolume destVolume; + do { + destVolume = hddsVolumes.get(rand.nextInt(hddsVolumes.size())); + } while (sourceVolume.equals(destVolume)); + + long containerSize = rand.nextInt(100 * 1024 * 1024) + 1; + + // Simulate deltaSizes update for the source volume(negative value) + deltaSizes.compute(sourceVolume, (k, v) -> (v == null ? 0L : v) - containerSize); + + // Simulate committedBytes update for the destination volume (space reserved) + destVolume.incCommittedBytes(containerSize); + } + + try { + Pair<HddsVolume, HddsVolume> pair = policy.chooseVolume(volumeSet, THRESHOLD, deltaSizes); + if (pair == null) { + volumeNotChosen++; + } else { + volumeChosen++; + // Note: In a real DiskBalancerService, after a successful choice, + // the committed bytes on the destination and delta on the source + // would be reverted/adjusted once the move completes or fails. + // This simulation focuses on the state before the policy makes a choice, + // assuming some background activity. For simplicity, we are not + // reverting these simulated incCommittedBytes or deltaSizes updates + // within this loop for each "chosen" pair. The random updates + // provide the dynamic background. + } + } catch (Exception e) { + failures++; + } + } + + long threadEnd = System.nanoTime(); + totalTimeNanos.addAndGet(threadEnd - threadStart); + pairChosenCount.addAndGet(volumeChosen); + pairNotChosenCount.addAndGet(volumeNotChosen); + failureCount.addAndGet(failures); + } finally { + latch.countDown(); + } + }); + } + + // Wait max 5 minutes for test completion + assertTrue(latch.await(5, TimeUnit.MINUTES), "Test timed out"); + + long totalOperations = (long) NUM_THREADS * NUM_ITERATIONS; + double avgTimePerOp = (double) totalTimeNanos.get() / totalOperations; + double opsPerSec = totalOperations / (totalTimeNanos.get() / 1_000_000_000.0); + + System.out.println("Performance results for " + policyName); + System.out.println("Total volumes: " + NUM_VOLUMES); + System.out.println("Total threads: " + NUM_THREADS); + System.out.println("Threshold(%): " + THRESHOLD * 100.0); + System.out.println("Total operations: " + totalOperations); + System.out.println("Volume Pair Chosen operations: " + pairChosenCount.get()); + System.out.println("Volume Pair Not Chosen operations: " + pairNotChosenCount.get()); + System.out.println("Failed operations: " + failureCount.get()); + System.out.println("Total time (ms): " + totalTimeNanos.get() / 1_000_000); + System.out.println("Average time per operation (ns): " + avgTimePerOp); + System.out.println("Operations per second: " + opsPerSec); + } + + private void initialiseVolumeSet() throws IOException { + // set a dummy path for initialisation, as we will override its internal volumeMap. + String initialDnDir = baseDir.resolve("initialDnDir").toFile().getAbsolutePath(); + Files.createDirectories(baseDir.resolve("initialDnDir")); + CONF.set(ScmConfigKeys.HDDS_DATANODE_DIR_KEY, initialDnDir); + + StateContext mockContext = mock(StateContext.class); + volumeSet = new MutableVolumeSet(UUID.randomUUID().toString(), CONF, + mockContext, StorageVolume.VolumeType.DATA_VOLUME, null); + } + + private void createVolumes() throws IOException { + // This map will replace the one inside 'volumeSet' + Map<String, StorageVolume> newVolumeMap = new ConcurrentHashMap<>(); + Random random = new Random(); + long startTime = System.currentTimeMillis(); + + for (int i = 0; i < NUM_VOLUMES; i++) { + String volumePath = baseDir.resolve("disk" + i).toString(); + //capacity of each volume is 1TB + long capacity = 1L * 1024 * 1024 * 1024 * 1024; + + // Distribute available space to create some variation in usage + // Between 5% and 75% Review Comment: 75% -> 55%? -- 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: issues-unsubscr...@ozone.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@ozone.apache.org For additional commands, e-mail: issues-h...@ozone.apache.org
