From f85ff03fa8eeb3ec92a8649b1679b1eb4b5df45c Mon Sep 17 00:00:00 2001 From: Rishikesh <62345295+Rishikesh1159@users.noreply.github.com> Date: Thu, 5 Sep 2024 12:31:21 -0700 Subject: [PATCH 1/7] [Backport 2.x] Add support for randomizing Remote Store enabled testing (#13845) * backport #12488 to 2.x branch. Signed-off-by: Rishikesh1159 <rishireddy1159@gmail.com> * Fix indentation. Signed-off-by: Rishikesh1159 <rishireddy1159@gmail.com> --------- Signed-off-by: Rishikesh1159 <rishireddy1159@gmail.com> Signed-off-by: Rishikesh <62345295+Rishikesh1159@users.noreply.github.com> --- .../recovery/ReplicaToPrimaryPromotionIT.java | 7 +++- .../test/OpenSearchIntegTestCase.java | 32 ++++++++++++++++--- 2 files changed, 34 insertions(+), 5 deletions(-) diff --git a/server/src/internalClusterTest/java/org/opensearch/indices/recovery/ReplicaToPrimaryPromotionIT.java b/server/src/internalClusterTest/java/org/opensearch/indices/recovery/ReplicaToPrimaryPromotionIT.java index 3df4ecff5250c..a2543f0592145 100644 --- a/server/src/internalClusterTest/java/org/opensearch/indices/recovery/ReplicaToPrimaryPromotionIT.java +++ b/server/src/internalClusterTest/java/org/opensearch/indices/recovery/ReplicaToPrimaryPromotionIT.java @@ -56,6 +56,11 @@ protected int numberOfReplicas() { return 1; } + @Override + public boolean useRandomReplicationStrategy() { + return true; + } + public void testPromoteReplicaToPrimary() throws Exception { final String indexName = randomAlphaOfLength(5).toLowerCase(Locale.ROOT); createIndex(indexName); @@ -65,7 +70,7 @@ public void testPromoteReplicaToPrimary() throws Exception { try (BackgroundIndexer indexer = new BackgroundIndexer(indexName, "_doc", client(), numOfDocs)) { waitForDocs(numOfDocs, indexer); } - refresh(indexName); + refreshAndWaitForReplication(indexName); } assertHitCount(client().prepareSearch(indexName).setSize(0).get(), numOfDocs); diff --git a/test/framework/src/main/java/org/opensearch/test/OpenSearchIntegTestCase.java b/test/framework/src/main/java/org/opensearch/test/OpenSearchIntegTestCase.java index 705610829cc9a..a63abdeb626b7 100644 --- a/test/framework/src/main/java/org/opensearch/test/OpenSearchIntegTestCase.java +++ b/test/framework/src/main/java/org/opensearch/test/OpenSearchIntegTestCase.java @@ -391,6 +391,14 @@ public abstract class OpenSearchIntegTestCase extends OpenSearchTestCase { */ public static final String TESTS_CLUSTER_NAME = "tests.clustername"; + protected static final String REMOTE_BACKED_STORAGE_REPOSITORY_NAME = "test-remote-store-repo"; + + private Path remoteStoreRepositoryPath; + + private ReplicationType randomReplicationType; + + private String randomStorageType; + /** * The lucene_default {@link Codec} is not added to the list as it internally maps to Asserting {@link Codec}. * The override to fetch the {@link CompletionFieldMapper.CompletionFieldType} postings format is not available for this codec. @@ -1985,11 +1993,19 @@ protected Settings nodeSettings(int nodeOrdinal) { builder.put(TelemetrySettings.TRACER_ENABLED_SETTING.getKey(), true); } - // Randomly set a replication strategy for the node. Replication Strategy can still be manually overridden by subclass if needed. + // Randomly set a Replication Strategy and storage type for the node. Both Replication Strategy and Storage Type can still be + // manually overridden by subclass if needed. if (useRandomReplicationStrategy()) { - ReplicationType replicationType = randomBoolean() ? ReplicationType.DOCUMENT : ReplicationType.SEGMENT; - logger.info("Randomly using Replication Strategy as {}.", replicationType.toString()); - builder.put(CLUSTER_REPLICATION_TYPE_SETTING.getKey(), replicationType); + if (randomReplicationType.equals(ReplicationType.SEGMENT) && randomStorageType.equals("REMOTE_STORE")) { + logger.info("Randomly using Replication Strategy as {} and Storage Type as {}.", randomReplicationType, randomStorageType); + if (remoteStoreRepositoryPath == null) { + remoteStoreRepositoryPath = randomRepoPath().toAbsolutePath(); + } + builder.put(remoteStoreClusterSettings(REMOTE_BACKED_STORAGE_REPOSITORY_NAME, remoteStoreRepositoryPath)); + } else { + logger.info("Randomly using Replication Strategy as {} and Storage Type as {}.", randomReplicationType, randomStorageType); + builder.put(CLUSTER_REPLICATION_TYPE_SETTING.getKey(), randomReplicationType); + } } return builder.build(); } @@ -2042,6 +2058,14 @@ protected boolean ignoreExternalCluster() { } protected TestCluster buildTestCluster(Scope scope, long seed) throws IOException { + if (useRandomReplicationStrategy()) { + randomReplicationType = randomBoolean() ? ReplicationType.DOCUMENT : ReplicationType.SEGMENT; + if (randomReplicationType.equals(ReplicationType.SEGMENT)) { + randomStorageType = randomBoolean() ? "REMOTE_STORE" : "LOCAL"; + } else { + randomStorageType = "LOCAL"; + } + } String clusterAddresses = System.getProperty(TESTS_CLUSTER); if (Strings.hasLength(clusterAddresses) && ignoreExternalCluster() == false) { if (scope == Scope.TEST) { From f2ecd3e763598e72c604cc1a0c616bae7dda1940 Mon Sep 17 00:00:00 2001 From: "opensearch-trigger-bot[bot]" <98922864+opensearch-trigger-bot[bot]@users.noreply.github.com> Date: Thu, 5 Sep 2024 16:06:03 -0400 Subject: [PATCH 2/7] Bump com.azure:azure-identity from 1.13.0 to 1.13.2 in /plugins/repository-azure (#15578) (#15755) * Bump com.azure:azure-identity in /plugins/repository-azure Bumps [com.azure:azure-identity](https://github.com/Azure/azure-sdk-for-java) from 1.13.0 to 1.13.2. - [Release notes](https://github.com/Azure/azure-sdk-for-java/releases) - [Commits](https://github.com/Azure/azure-sdk-for-java/compare/azure-core_1.13.0...azure-identity_1.13.2) --- updated-dependencies: - dependency-name: com.azure:azure-identity dependency-type: direct:production update-type: version-update:semver-patch ... * Updating SHAs * Update changelog --------- (cherry picked from commit e5e8504acdf1339f53dced4c00f5c9cd91b95f2c) Signed-off-by: dependabot[bot] <support@github.com> Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: dependabot[bot] <dependabot[bot]@users.noreply.github.com> --- CHANGELOG.md | 1 + plugins/repository-azure/build.gradle | 2 +- .../repository-azure/licenses/azure-identity-1.13.0.jar.sha1 | 1 - .../repository-azure/licenses/azure-identity-1.13.2.jar.sha1 | 1 + 4 files changed, 3 insertions(+), 2 deletions(-) delete mode 100644 plugins/repository-azure/licenses/azure-identity-1.13.0.jar.sha1 create mode 100644 plugins/repository-azure/licenses/azure-identity-1.13.2.jar.sha1 diff --git a/CHANGELOG.md b/CHANGELOG.md index 5a8391dd79d5f..4dbc667fcdbbf 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -78,6 +78,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), - Bump `com.netflix.nebula.ospackage-base` from 11.9.1 to 11.10.0 ([#15419](https://github.com/opensearch-project/OpenSearch/pull/15419)) - Bump `org.roaringbitmap:RoaringBitmap` from 1.1.0 to 1.2.1 ([#15423](https://github.com/opensearch-project/OpenSearch/pull/15423)) - Bump `icu4j` from 70.1 to 75.1 ([#15469](https://github.com/opensearch-project/OpenSearch/pull/15469)) +- Bump `com.azure:azure-identity` from 1.13.0 to 1.13.2 ([#15578](https://github.com/opensearch-project/OpenSearch/pull/15578)) ### Changed - Add lower limit for primary and replica batch allocators timeout ([#14979](https://github.com/opensearch-project/OpenSearch/pull/14979)) diff --git a/plugins/repository-azure/build.gradle b/plugins/repository-azure/build.gradle index e5d3bd215f8cb..e3dd8c1bd0dfa 100644 --- a/plugins/repository-azure/build.gradle +++ b/plugins/repository-azure/build.gradle @@ -57,7 +57,7 @@ dependencies { api "io.netty:netty-transport-native-unix-common:${versions.netty}" implementation project(':modules:transport-netty4') api 'com.azure:azure-storage-blob:12.23.0' - api 'com.azure:azure-identity:1.13.0' + api 'com.azure:azure-identity:1.13.2' // Start of transitive dependencies for azure-identity api 'com.microsoft.azure:msal4j-persistence-extension:1.3.0' api "net.java.dev.jna:jna-platform:${versions.jna}" diff --git a/plugins/repository-azure/licenses/azure-identity-1.13.0.jar.sha1 b/plugins/repository-azure/licenses/azure-identity-1.13.0.jar.sha1 deleted file mode 100644 index b59c2a3be5c92..0000000000000 --- a/plugins/repository-azure/licenses/azure-identity-1.13.0.jar.sha1 +++ /dev/null @@ -1 +0,0 @@ -54b44a74636322d06e9dc42d611a9f12a0966790 \ No newline at end of file diff --git a/plugins/repository-azure/licenses/azure-identity-1.13.2.jar.sha1 b/plugins/repository-azure/licenses/azure-identity-1.13.2.jar.sha1 new file mode 100644 index 0000000000000..7c98a9ccba592 --- /dev/null +++ b/plugins/repository-azure/licenses/azure-identity-1.13.2.jar.sha1 @@ -0,0 +1 @@ +50a1daef3eb5c6ab2e1351a3e3f5a7649a8fe464 \ No newline at end of file From d5e86e13ec950fda97e99bd75b1e34d99e10fd45 Mon Sep 17 00:00:00 2001 From: Arpit-Bandejiya <abandeji@amazon.com> Date: Fri, 6 Sep 2024 01:37:24 +0530 Subject: [PATCH 3/7] Schedule reroute after allocator timed out (#15565) (#15749) * Schedule reroute after allocator timed out Signed-off-by: Rishab Nahata <rnnahata@amazon.com> (cherry picked from commit 4f50b4d705deaa971d802cd22fb120f75c722517) Co-authored-by: Rishab Nahata <rnnahata@amazon.com> --- .../org/opensearch/cluster/ClusterModule.java | 4 + .../allocator/BalancedShardsAllocator.java | 25 ++ .../allocation/allocator/ShardsAllocator.java | 3 + .../gateway/ShardsBatchGatewayAllocator.java | 30 +- .../main/java/org/opensearch/node/Node.java | 1 + .../cluster/ClusterModuleTests.java | 13 + ...TimeBoundBalancedShardsAllocatorTests.java | 257 +++++++++++++----- .../gateway/GatewayAllocatorTests.java | 62 ++++- .../TestShardBatchGatewayAllocator.java | 6 +- 9 files changed, 321 insertions(+), 80 deletions(-) diff --git a/server/src/main/java/org/opensearch/cluster/ClusterModule.java b/server/src/main/java/org/opensearch/cluster/ClusterModule.java index 0d15158f31e34..9e2074c7aacf9 100644 --- a/server/src/main/java/org/opensearch/cluster/ClusterModule.java +++ b/server/src/main/java/org/opensearch/cluster/ClusterModule.java @@ -52,6 +52,7 @@ import org.opensearch.cluster.metadata.RepositoriesMetadata; import org.opensearch.cluster.metadata.WeightedRoutingMetadata; import org.opensearch.cluster.routing.DelayedAllocationService; +import org.opensearch.cluster.routing.RerouteService; import org.opensearch.cluster.routing.allocation.AllocationService; import org.opensearch.cluster.routing.allocation.ExistingShardsAllocator; import org.opensearch.cluster.routing.allocation.allocator.BalancedShardsAllocator; @@ -476,4 +477,7 @@ public void setExistingShardsAllocators(GatewayAllocator gatewayAllocator, Shard allocationService.setExistingShardsAllocators(existingShardsAllocators); } + public void setRerouteServiceForAllocator(RerouteService rerouteService) { + shardsAllocator.setRerouteService(rerouteService); + } } diff --git a/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/BalancedShardsAllocator.java b/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/BalancedShardsAllocator.java index a5193ca602f04..785636fa7ff2a 100644 --- a/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/BalancedShardsAllocator.java +++ b/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/BalancedShardsAllocator.java @@ -35,6 +35,7 @@ import org.apache.logging.log4j.LogManager; import org.apache.logging.log4j.Logger; import org.apache.lucene.util.IntroSorter; +import org.opensearch.cluster.routing.RerouteService; import org.opensearch.cluster.routing.RoutingNode; import org.opensearch.cluster.routing.RoutingNodes; import org.opensearch.cluster.routing.ShardMovementStrategy; @@ -49,12 +50,14 @@ import org.opensearch.cluster.routing.allocation.RebalanceParameter; import org.opensearch.cluster.routing.allocation.RoutingAllocation; import org.opensearch.cluster.routing.allocation.ShardAllocationDecision; +import org.opensearch.common.Priority; import org.opensearch.common.inject.Inject; import org.opensearch.common.settings.ClusterSettings; import org.opensearch.common.settings.Setting; import org.opensearch.common.settings.Setting.Property; import org.opensearch.common.settings.Settings; import org.opensearch.common.unit.TimeValue; +import org.opensearch.core.action.ActionListener; import java.util.HashMap; import java.util.HashSet; @@ -202,6 +205,7 @@ public class BalancedShardsAllocator implements ShardsAllocator { private volatile boolean ignoreThrottleInRestore; private volatile TimeValue allocatorTimeout; private long startTime; + private RerouteService rerouteService; public BalancedShardsAllocator(Settings settings) { this(settings, new ClusterSettings(settings, ClusterSettings.BUILT_IN_CLUSTER_SETTINGS)); @@ -231,6 +235,12 @@ public BalancedShardsAllocator(Settings settings, ClusterSettings clusterSetting clusterSettings.addSettingsUpdateConsumer(ALLOCATOR_TIMEOUT_SETTING, this::setAllocatorTimeout); } + @Override + public void setRerouteService(RerouteService rerouteService) { + assert this.rerouteService == null : "RerouteService is already set"; + this.rerouteService = rerouteService; + } + /** * Changes in deprecated setting SHARD_MOVE_PRIMARY_FIRST_SETTING affect value of its replacement setting SHARD_MOVEMENT_STRATEGY_SETTING. */ @@ -342,6 +352,7 @@ public void allocate(RoutingAllocation allocation) { localShardsBalancer.allocateUnassigned(); localShardsBalancer.moveShards(); localShardsBalancer.balance(); + scheduleRerouteIfAllocatorTimedOut(); final ShardsBalancer remoteShardsBalancer = new RemoteShardsBalancer(logger, allocation); remoteShardsBalancer.allocateUnassigned(); @@ -404,6 +415,20 @@ private void failAllocationOfNewPrimaries(RoutingAllocation allocation) { } } + private void scheduleRerouteIfAllocatorTimedOut() { + if (allocatorTimedOut()) { + assert rerouteService != null : "RerouteService not set to schedule reroute after allocator time out"; + rerouteService.reroute( + "reroute after balanced shards allocator timed out", + Priority.HIGH, + ActionListener.wrap( + r -> logger.trace("reroute after balanced shards allocator timed out completed"), + e -> logger.debug("reroute after balanced shards allocator timed out failed", e) + ) + ); + } + } + /** * Returns the currently configured delta threshold */ diff --git a/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/ShardsAllocator.java b/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/ShardsAllocator.java index 29e9acca4e6c2..38aafff6ce3e8 100644 --- a/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/ShardsAllocator.java +++ b/server/src/main/java/org/opensearch/cluster/routing/allocation/allocator/ShardsAllocator.java @@ -32,6 +32,7 @@ package org.opensearch.cluster.routing.allocation.allocator; +import org.opensearch.cluster.routing.RerouteService; import org.opensearch.cluster.routing.ShardRouting; import org.opensearch.cluster.routing.allocation.AllocateUnassignedDecision; import org.opensearch.cluster.routing.allocation.MoveDecision; @@ -73,4 +74,6 @@ public interface ShardsAllocator { * the cluster explain API, then this method should throw a {@code UnsupportedOperationException}. */ ShardAllocationDecision decideShardAllocation(ShardRouting shard, RoutingAllocation allocation); + + default void setRerouteService(RerouteService rerouteService) {} } diff --git a/server/src/main/java/org/opensearch/gateway/ShardsBatchGatewayAllocator.java b/server/src/main/java/org/opensearch/gateway/ShardsBatchGatewayAllocator.java index d18304ea73ed0..5e2dcbcd70b40 100644 --- a/server/src/main/java/org/opensearch/gateway/ShardsBatchGatewayAllocator.java +++ b/server/src/main/java/org/opensearch/gateway/ShardsBatchGatewayAllocator.java @@ -184,11 +184,11 @@ public void cleanCaches() { // for tests protected ShardsBatchGatewayAllocator() { - this(DEFAULT_SHARD_BATCH_SIZE); + this(DEFAULT_SHARD_BATCH_SIZE, null); } - protected ShardsBatchGatewayAllocator(long batchSize) { - this.rerouteService = null; + protected ShardsBatchGatewayAllocator(long batchSize, RerouteService rerouteService) { + this.rerouteService = rerouteService; this.batchStartedAction = null; this.primaryShardBatchAllocator = null; this.batchStoreAction = null; @@ -297,6 +297,18 @@ public void run() { public void onComplete() { logger.trace("Triggering oncomplete after timeout for [{}] primary shards", timedOutPrimaryShardIds.size()); primaryBatchShardAllocator.allocateUnassignedBatchOnTimeout(timedOutPrimaryShardIds, allocation, true); + if (timedOutPrimaryShardIds.isEmpty() == false) { + logger.trace("scheduling reroute after existing shards allocator timed out for primary shards"); + assert rerouteService != null; + rerouteService.reroute( + "reroute after existing shards allocator timed out", + Priority.HIGH, + ActionListener.wrap( + r -> logger.trace("reroute after existing shards allocator timed out completed"), + e -> logger.debug("reroute after existing shards allocator timed out failed", e) + ) + ); + } } }; } else { @@ -320,6 +332,18 @@ public void run() { public void onComplete() { logger.trace("Triggering oncomplete after timeout for [{}] replica shards", timedOutReplicaShardIds.size()); replicaBatchShardAllocator.allocateUnassignedBatchOnTimeout(timedOutReplicaShardIds, allocation, false); + if (timedOutReplicaShardIds.isEmpty() == false) { + logger.trace("scheduling reroute after existing shards allocator timed out for replica shards"); + assert rerouteService != null; + rerouteService.reroute( + "reroute after existing shards allocator timed out", + Priority.HIGH, + ActionListener.wrap( + r -> logger.trace("reroute after existing shards allocator timed out completed"), + e -> logger.debug("reroute after existing shards allocator timed out failed", e) + ) + ); + } } }; } diff --git a/server/src/main/java/org/opensearch/node/Node.java b/server/src/main/java/org/opensearch/node/Node.java index 26150bdecdcf5..72531b98ec102 100644 --- a/server/src/main/java/org/opensearch/node/Node.java +++ b/server/src/main/java/org/opensearch/node/Node.java @@ -867,6 +867,7 @@ protected Node( final RerouteService rerouteService = new BatchedRerouteService(clusterService, clusterModule.getAllocationService()::reroute); rerouteServiceReference.set(rerouteService); clusterService.setRerouteService(rerouteService); + clusterModule.setRerouteServiceForAllocator(rerouteService); final RecoverySettings recoverySettings = new RecoverySettings(settings, settingsModule.getClusterSettings()); diff --git a/server/src/test/java/org/opensearch/cluster/ClusterModuleTests.java b/server/src/test/java/org/opensearch/cluster/ClusterModuleTests.java index 97706927ba857..f8240e775cfa5 100644 --- a/server/src/test/java/org/opensearch/cluster/ClusterModuleTests.java +++ b/server/src/test/java/org/opensearch/cluster/ClusterModuleTests.java @@ -337,6 +337,19 @@ public void testQueryGroupMetadataRegister() { ); } + public void testRerouteServiceSetForBalancedShardsAllocator() { + ClusterModule clusterModule = new ClusterModule( + Settings.EMPTY, + clusterService, + Collections.emptyList(), + clusterInfoService, + null, + threadContext, + new ClusterManagerMetrics(NoopMetricsRegistry.INSTANCE) + ); + clusterModule.setRerouteServiceForAllocator((reason, priority, listener) -> listener.onResponse(clusterService.state())); + } + private static ClusterPlugin existingShardsAllocatorPlugin(final String allocatorName) { return new ClusterPlugin() { @Override diff --git a/server/src/test/java/org/opensearch/cluster/routing/allocation/allocator/TimeBoundBalancedShardsAllocatorTests.java b/server/src/test/java/org/opensearch/cluster/routing/allocation/allocator/TimeBoundBalancedShardsAllocatorTests.java index a10c305686638..45a0bd7b18afd 100644 --- a/server/src/test/java/org/opensearch/cluster/routing/allocation/allocator/TimeBoundBalancedShardsAllocatorTests.java +++ b/server/src/test/java/org/opensearch/cluster/routing/allocation/allocator/TimeBoundBalancedShardsAllocatorTests.java @@ -8,6 +8,7 @@ package org.opensearch.cluster.routing.allocation.allocator; +import org.opensearch.OpenSearchException; import org.opensearch.Version; import org.opensearch.cluster.ClusterInfo; import org.opensearch.cluster.ClusterName; @@ -17,6 +18,7 @@ import org.opensearch.cluster.metadata.Metadata; import org.opensearch.cluster.node.DiscoveryNode; import org.opensearch.cluster.node.DiscoveryNodes; +import org.opensearch.cluster.routing.RerouteService; import org.opensearch.cluster.routing.RoutingNodes; import org.opensearch.cluster.routing.RoutingTable; import org.opensearch.cluster.routing.ShardRouting; @@ -26,14 +28,20 @@ import org.opensearch.cluster.routing.allocation.decider.AllocationDeciders; import org.opensearch.cluster.routing.allocation.decider.Decision; import org.opensearch.cluster.routing.allocation.decider.SameShardAllocationDecider; +import org.opensearch.cluster.service.ClusterService; +import org.opensearch.common.Priority; import org.opensearch.common.settings.ClusterSettings; import org.opensearch.common.settings.Settings; +import org.opensearch.test.ClusterServiceUtils; +import org.opensearch.threadpool.TestThreadPool; +import org.junit.After; import java.util.Arrays; import java.util.Collections; import java.util.List; import java.util.Map; import java.util.concurrent.CountDownLatch; +import java.util.concurrent.atomic.AtomicBoolean; import static org.opensearch.cluster.routing.ShardRoutingState.INITIALIZING; import static org.opensearch.cluster.routing.ShardRoutingState.STARTED; @@ -41,26 +49,49 @@ public class TimeBoundBalancedShardsAllocatorTests extends OpenSearchAllocationTestCase { + private TestThreadPool threadPool; + private ClusterService clusterService; + private ClusterState state; + private final DiscoveryNode node1 = newNode("node1", "node1", Collections.singletonMap("zone", "1a")); private final DiscoveryNode node2 = newNode("node2", "node2", Collections.singletonMap("zone", "1b")); private final DiscoveryNode node3 = newNode("node3", "node3", Collections.singletonMap("zone", "1c")); - public void testAllUnassignedShardsAllocatedWhenNoTimeOut() { + @After + @Override + public void tearDown() throws Exception { + super.tearDown(); + if (threadPool != null) { + final boolean terminated = terminate(threadPool); + assert terminated; + } + if (clusterService != null) { + clusterService.close(); + } + } + + public void setupStateAndService(Metadata metadata, RoutingTable routingTable) { + state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) + .metadata(metadata) + .routingTable(routingTable) + .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) + .build(); + threadPool = new TestThreadPool(getTestName()); + clusterService = ClusterServiceUtils.createClusterService(state, threadPool); + } + + public void testAllUnassignedShardsAllocatedWhenNoTimeOutAndRerouteNotScheduled() { int numberOfIndices = 2; int numberOfShards = 5; int numberOfReplicas = 1; int totalPrimaryCount = numberOfIndices * numberOfShards; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Settings.Builder settings = Settings.builder(); - // passing total shard count for timed out latch such that no shard times out - BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings.build(), new CountDownLatch(totalShardCount)); + // passing sufficiently high count for timeout latch to simulate no time out + BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings.build(), new CountDownLatch(Integer.MAX_VALUE)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); RoutingAllocation allocation = new RoutingAllocation( yesAllocationDeciders(), new RoutingNodes(state, false), @@ -69,6 +100,18 @@ public void testAllUnassignedShardsAllocatedWhenNoTimeOut() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> initializingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.INITIALIZING); int node1Recoveries = allocation.routingNodes().getInitialPrimariesIncomingRecoveries(node1.getId()); @@ -77,9 +120,10 @@ public void testAllUnassignedShardsAllocatedWhenNoTimeOut() { assertEquals(totalShardCount, initializingShards.size()); assertEquals(0, allocation.routingNodes().unassigned().ignored().size()); assertEquals(totalPrimaryCount, node1Recoveries + node2Recoveries + node3Recoveries); + assertFalse(rerouteScheduled.get()); } - public void testAllUnassignedShardsIgnoredWhenTimedOut() { + public void testAllUnassignedShardsIgnoredWhenTimedOutAndRerouteScheduled() { int numberOfIndices = 2; int numberOfShards = 5; int numberOfReplicas = 1; @@ -89,11 +133,7 @@ public void testAllUnassignedShardsIgnoredWhenTimedOut() { BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings.build(), new CountDownLatch(0)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); RoutingAllocation allocation = new RoutingAllocation( yesAllocationDeciders(), new RoutingNodes(state, false), @@ -102,6 +142,18 @@ public void testAllUnassignedShardsIgnoredWhenTimedOut() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> initializingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.INITIALIZING); int node1Recoveries = allocation.routingNodes().getInitialPrimariesIncomingRecoveries(node1.getId()); @@ -110,9 +162,10 @@ public void testAllUnassignedShardsIgnoredWhenTimedOut() { assertEquals(0, initializingShards.size()); assertEquals(totalShardCount, allocation.routingNodes().unassigned().ignored().size()); assertEquals(0, node1Recoveries + node2Recoveries + node3Recoveries); + assertTrue(rerouteScheduled.get()); } - public void testAllocatePartialPrimaryShardsUntilTimedOut() { + public void testAllocatePartialPrimaryShardsUntilTimedOutAndRerouteScheduled() { int numberOfIndices = 2; int numberOfShards = 5; int numberOfReplicas = 1; @@ -123,11 +176,7 @@ public void testAllocatePartialPrimaryShardsUntilTimedOut() { BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings.build(), new CountDownLatch(shardsToAllocate)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); RoutingAllocation allocation = new RoutingAllocation( yesAllocationDeciders(), new RoutingNodes(state, false), @@ -136,6 +185,18 @@ public void testAllocatePartialPrimaryShardsUntilTimedOut() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> initializingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.INITIALIZING); int node1Recoveries = allocation.routingNodes().getInitialPrimariesIncomingRecoveries(node1.getId()); @@ -144,9 +205,10 @@ public void testAllocatePartialPrimaryShardsUntilTimedOut() { assertEquals(shardsToAllocate, initializingShards.size()); assertEquals(totalShardCount - shardsToAllocate, allocation.routingNodes().unassigned().ignored().size()); assertEquals(shardsToAllocate, node1Recoveries + node2Recoveries + node3Recoveries); + assertTrue(rerouteScheduled.get()); } - public void testAllocateAllPrimaryShardsAndPartialReplicaShardsUntilTimedOut() { + public void testAllocateAllPrimaryShardsAndPartialReplicaShardsUntilTimedOutAndRerouteScheduled() { int numberOfIndices = 2; int numberOfShards = 5; int numberOfReplicas = 1; @@ -158,11 +220,7 @@ public void testAllocateAllPrimaryShardsAndPartialReplicaShardsUntilTimedOut() { BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings.build(), new CountDownLatch(shardsToAllocate)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); RoutingAllocation allocation = new RoutingAllocation( yesAllocationDeciders(), new RoutingNodes(state, false), @@ -171,6 +229,18 @@ public void testAllocateAllPrimaryShardsAndPartialReplicaShardsUntilTimedOut() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> initializingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.INITIALIZING); int node1Recoveries = allocation.routingNodes().getInitialPrimariesIncomingRecoveries(node1.getId()); @@ -179,20 +249,17 @@ public void testAllocateAllPrimaryShardsAndPartialReplicaShardsUntilTimedOut() { assertEquals(shardsToAllocate, initializingShards.size()); assertEquals(totalShardCount - shardsToAllocate, allocation.routingNodes().unassigned().ignored().size()); assertEquals(numberOfShards * numberOfIndices, node1Recoveries + node2Recoveries + node3Recoveries); + assertTrue(rerouteScheduled.get()); } - public void testAllShardsMoveWhenExcludedAndTimeoutNotBreached() { + public void testAllShardsMoveWhenExcludedAndTimeoutNotBreachedAndRerouteNotScheduled() { int numberOfIndices = 3; int numberOfShards = 5; int numberOfReplicas = 1; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); MockAllocationService allocationService = createAllocationService(); state = applyStartedShardsUntilNoChange(state, allocationService); // check all shards allocated @@ -200,8 +267,7 @@ public void testAllShardsMoveWhenExcludedAndTimeoutNotBreached() { assertEquals(totalShardCount, state.getRoutingNodes().shardsWithState(STARTED).size()); int node1ShardCount = state.getRoutingNodes().node("node1").size(); Settings settings = Settings.builder().put("cluster.routing.allocation.exclude.zone", "1a").build(); - int shardsToMove = 10 + 1000; // such that time out is never breached - BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings, new CountDownLatch(shardsToMove)); + BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings, new CountDownLatch(Integer.MAX_VALUE)); RoutingAllocation allocation = new RoutingAllocation( allocationDecidersForExcludeAPI(settings), new RoutingNodes(state, false), @@ -210,30 +276,39 @@ public void testAllShardsMoveWhenExcludedAndTimeoutNotBreached() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> relocatingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.RELOCATING); assertEquals(node1ShardCount, relocatingShards.size()); + assertFalse(rerouteScheduled.get()); } - public void testNoShardsMoveWhenExcludedAndTimeoutBreached() { + public void testNoShardsMoveWhenExcludedAndTimeoutBreachedAndRerouteScheduled() { int numberOfIndices = 3; int numberOfShards = 5; int numberOfReplicas = 1; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); MockAllocationService allocationService = createAllocationService(); state = applyStartedShardsUntilNoChange(state, allocationService); // check all shards allocated assertEquals(0, state.getRoutingNodes().shardsWithState(INITIALIZING).size()); assertEquals(totalShardCount, state.getRoutingNodes().shardsWithState(STARTED).size()); Settings settings = Settings.builder().put("cluster.routing.allocation.exclude.zone", "1a").build(); - int shardsToMove = 0; // such that time out is never breached + int shardsToMove = 0; // such that time out is breached BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(settings, new CountDownLatch(shardsToMove)); RoutingAllocation allocation = new RoutingAllocation( allocationDecidersForExcludeAPI(settings), @@ -243,23 +318,32 @@ public void testNoShardsMoveWhenExcludedAndTimeoutBreached() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> relocatingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.RELOCATING); assertEquals(0, relocatingShards.size()); + assertTrue(rerouteScheduled.get()); } - public void testPartialShardsMoveWhenExcludedAndTimeoutBreached() { + public void testPartialShardsMoveWhenExcludedAndTimeoutBreachedAndRerouteScheduled() { int numberOfIndices = 3; int numberOfShards = 5; int numberOfReplicas = 1; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); MockAllocationService allocationService = createAllocationService(); state = applyStartedShardsUntilNoChange(state, allocationService); // check all shards allocated @@ -279,23 +363,32 @@ public void testPartialShardsMoveWhenExcludedAndTimeoutBreached() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> relocatingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.RELOCATING); assertEquals(shardsToMove / 3, relocatingShards.size()); + assertTrue(rerouteScheduled.get()); } - public void testClusterRebalancedWhenNotTimedOut() { + public void testClusterRebalancedWhenNotTimedOutAndRerouteNotScheduled() { int numberOfIndices = 1; int numberOfShards = 15; int numberOfReplicas = 1; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); MockAllocationService allocationService = createAllocationService( Settings.builder().put("cluster.routing.allocation.exclude.zone", "1a").build() ); // such that no shards are allocated to node1 @@ -306,8 +399,7 @@ public void testClusterRebalancedWhenNotTimedOut() { assertEquals(totalShardCount, state.getRoutingNodes().shardsWithState(STARTED).size()); assertEquals(0, node1ShardCount); Settings newSettings = Settings.builder().put("cluster.routing.allocation.exclude.zone", "").build(); - int shardsToMove = 1000; // such that time out is never breached - BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(newSettings, new CountDownLatch(shardsToMove)); + BalancedShardsAllocator allocator = new TestBalancedShardsAllocator(newSettings, new CountDownLatch(Integer.MAX_VALUE)); RoutingAllocation allocation = new RoutingAllocation( allocationDecidersForExcludeAPI(newSettings), new RoutingNodes(state, false), @@ -316,23 +408,32 @@ public void testClusterRebalancedWhenNotTimedOut() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> relocatingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.RELOCATING); assertEquals(totalShardCount / 3, relocatingShards.size()); + assertFalse(rerouteScheduled.get()); } - public void testClusterNotRebalancedWhenTimedOut() { + public void testClusterNotRebalancedWhenTimedOutAndRerouteScheduled() { int numberOfIndices = 1; int numberOfShards = 15; int numberOfReplicas = 1; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); MockAllocationService allocationService = createAllocationService( Settings.builder().put("cluster.routing.allocation.exclude.zone", "1a").build() ); // such that no shards are allocated to node1 @@ -353,23 +454,32 @@ public void testClusterNotRebalancedWhenTimedOut() { null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> relocatingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.RELOCATING); assertEquals(0, relocatingShards.size()); + assertTrue(rerouteScheduled.get()); } - public void testClusterPartialRebalancedWhenTimedOut() { + public void testClusterPartialRebalancedWhenTimedOutAndRerouteScheduled() { int numberOfIndices = 1; int numberOfShards = 15; int numberOfReplicas = 1; int totalShardCount = numberOfIndices * (numberOfShards * (numberOfReplicas + 1)); Metadata metadata = buildMetadata(Metadata.builder(), numberOfIndices, numberOfShards, numberOfReplicas); RoutingTable routingTable = buildRoutingTable(metadata); - ClusterState state = ClusterState.builder(ClusterName.CLUSTER_NAME_SETTING.getDefault(Settings.EMPTY)) - .metadata(metadata) - .routingTable(routingTable) - .nodes(DiscoveryNodes.builder().add(node1).add(node2).add(node3)) - .build(); + setupStateAndService(metadata, routingTable); MockAllocationService allocationService = createAllocationService( Settings.builder().put("cluster.routing.allocation.exclude.zone", "1a").build() ); // such that no shards are allocated to node1 @@ -404,9 +514,22 @@ public Decision canRebalance(ShardRouting shardRouting, RoutingAllocation alloca null, System.nanoTime() ); + AtomicBoolean rerouteScheduled = new AtomicBoolean(false); + final RerouteService rerouteService = (reason, priority, listener) -> { + if (randomBoolean()) { + listener.onFailure(new OpenSearchException("simulated")); + } else { + listener.onResponse(clusterService.state()); + } + assertEquals("reroute after balanced shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteScheduled.compareAndSet(false, true); + }; + allocator.setRerouteService(rerouteService); allocator.allocate(allocation); List<ShardRouting> relocatingShards = allocation.routingNodes().shardsWithState(ShardRoutingState.RELOCATING); assertEquals(3, relocatingShards.size()); + assertTrue(rerouteScheduled.get()); } public void testAllocatorNeverTimedOutIfValueIsMinusOne() { diff --git a/server/src/test/java/org/opensearch/gateway/GatewayAllocatorTests.java b/server/src/test/java/org/opensearch/gateway/GatewayAllocatorTests.java index c7eae77d6deba..ebc2e59fa5a30 100644 --- a/server/src/test/java/org/opensearch/gateway/GatewayAllocatorTests.java +++ b/server/src/test/java/org/opensearch/gateway/GatewayAllocatorTests.java @@ -10,6 +10,7 @@ import org.apache.logging.log4j.LogManager; import org.apache.logging.log4j.Logger; +import org.opensearch.OpenSearchException; import org.opensearch.Version; import org.opensearch.action.support.nodes.BaseNodeResponse; import org.opensearch.cluster.ClusterInfo; @@ -22,6 +23,7 @@ import org.opensearch.cluster.routing.IndexRoutingTable; import org.opensearch.cluster.routing.IndexShardRoutingTable; import org.opensearch.cluster.routing.RecoverySource; +import org.opensearch.cluster.routing.RerouteService; import org.opensearch.cluster.routing.RoutingNodes; import org.opensearch.cluster.routing.RoutingTable; import org.opensearch.cluster.routing.ShardRouting; @@ -30,6 +32,8 @@ import org.opensearch.cluster.routing.UnassignedInfo; import org.opensearch.cluster.routing.allocation.RoutingAllocation; import org.opensearch.cluster.routing.allocation.decider.AllocationDeciders; +import org.opensearch.cluster.service.ClusterService; +import org.opensearch.common.Priority; import org.opensearch.common.collect.Tuple; import org.opensearch.common.settings.Settings; import org.opensearch.common.unit.TimeValue; @@ -37,7 +41,9 @@ import org.opensearch.common.util.set.Sets; import org.opensearch.core.index.shard.ShardId; import org.opensearch.snapshots.SnapshotShardSizeInfo; +import org.opensearch.test.ClusterServiceUtils; import org.opensearch.test.gateway.TestShardBatchGatewayAllocator; +import org.opensearch.threadpool.TestThreadPool; import org.junit.Before; import java.util.ArrayList; @@ -47,13 +53,13 @@ import java.util.Map; import java.util.Set; import java.util.concurrent.CountDownLatch; -import java.util.concurrent.TimeUnit; import java.util.stream.Collectors; import static org.opensearch.gateway.ShardsBatchGatewayAllocator.PRIMARY_BATCH_ALLOCATOR_TIMEOUT_SETTING; import static org.opensearch.gateway.ShardsBatchGatewayAllocator.PRIMARY_BATCH_ALLOCATOR_TIMEOUT_SETTING_KEY; import static org.opensearch.gateway.ShardsBatchGatewayAllocator.REPLICA_BATCH_ALLOCATOR_TIMEOUT_SETTING; import static org.opensearch.gateway.ShardsBatchGatewayAllocator.REPLICA_BATCH_ALLOCATOR_TIMEOUT_SETTING_KEY; +import static org.hamcrest.Matchers.greaterThanOrEqualTo; public class GatewayAllocatorTests extends OpenSearchAllocationTestCase { @@ -426,22 +432,62 @@ public void testReplicaAllocatorTimeout() { assertEquals(-1, REPLICA_BATCH_ALLOCATOR_TIMEOUT_SETTING.get(build).getMillis()); } - public void testCollectTimedOutShards() throws InterruptedException { + public void testCollectTimedOutShardsAndScheduleReroute_Success() throws InterruptedException { createIndexAndUpdateClusterState(2, 5, 2); - CountDownLatch latch = new CountDownLatch(10); - testShardsBatchGatewayAllocator = new TestShardBatchGatewayAllocator(latch); + TestThreadPool threadPool = new TestThreadPool(getTestName()); + ClusterService clusterService = ClusterServiceUtils.createClusterService(clusterState, threadPool); + final CountDownLatch rerouteLatch = new CountDownLatch(2); + final RerouteService rerouteService = (reason, priority, listener) -> { + listener.onResponse(clusterService.state()); + assertThat(rerouteLatch.getCount(), greaterThanOrEqualTo(0L)); + assertEquals("reroute after existing shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteLatch.countDown(); + }; + CountDownLatch timedOutShardsLatch = new CountDownLatch(20); + testShardsBatchGatewayAllocator = new TestShardBatchGatewayAllocator(timedOutShardsLatch, 1000, rerouteService); testShardsBatchGatewayAllocator.setPrimaryBatchAllocatorTimeout(TimeValue.ZERO); testShardsBatchGatewayAllocator.setReplicaBatchAllocatorTimeout(TimeValue.ZERO); BatchRunnableExecutor executor = testShardsBatchGatewayAllocator.allocateAllUnassignedShards(testAllocation, true); executor.run(); - assertTrue(latch.await(1, TimeUnit.MINUTES)); - latch = new CountDownLatch(10); - testShardsBatchGatewayAllocator = new TestShardBatchGatewayAllocator(latch); + assertEquals(timedOutShardsLatch.getCount(), 10); + assertEquals(1, rerouteLatch.getCount()); + executor = testShardsBatchGatewayAllocator.allocateAllUnassignedShards(testAllocation, false); + executor.run(); + assertEquals(timedOutShardsLatch.getCount(), 0); + assertEquals(0, rerouteLatch.getCount()); // even with failure it doesn't leak any listeners + final boolean terminated = terminate(threadPool); + assert terminated; + clusterService.close(); + } + + public void testCollectTimedOutShardsAndScheduleReroute_Failure() throws InterruptedException { + createIndexAndUpdateClusterState(2, 5, 2); + TestThreadPool threadPool = new TestThreadPool(getTestName()); + ClusterService clusterService = ClusterServiceUtils.createClusterService(clusterState, threadPool); + final CountDownLatch rerouteLatch = new CountDownLatch(2); + final RerouteService rerouteService = (reason, priority, listener) -> { + listener.onFailure(new OpenSearchException("simulated")); + assertThat(rerouteLatch.getCount(), greaterThanOrEqualTo(0L)); + assertEquals("reroute after existing shards allocator timed out", reason); + assertEquals(Priority.HIGH, priority); + rerouteLatch.countDown(); + }; + CountDownLatch timedOutShardsLatch = new CountDownLatch(20); + testShardsBatchGatewayAllocator = new TestShardBatchGatewayAllocator(timedOutShardsLatch, 1000, rerouteService); testShardsBatchGatewayAllocator.setPrimaryBatchAllocatorTimeout(TimeValue.ZERO); testShardsBatchGatewayAllocator.setReplicaBatchAllocatorTimeout(TimeValue.ZERO); + BatchRunnableExecutor executor = testShardsBatchGatewayAllocator.allocateAllUnassignedShards(testAllocation, true); + executor.run(); + assertEquals(timedOutShardsLatch.getCount(), 10); + assertEquals(1, rerouteLatch.getCount()); executor = testShardsBatchGatewayAllocator.allocateAllUnassignedShards(testAllocation, false); executor.run(); - assertTrue(latch.await(1, TimeUnit.MINUTES)); + assertEquals(timedOutShardsLatch.getCount(), 0); + assertEquals(0, rerouteLatch.getCount()); // even with failure it doesn't leak any listeners + final boolean terminated = terminate(threadPool); + assert terminated; + clusterService.close(); } private void createIndexAndUpdateClusterState(int count, int numberOfShards, int numberOfReplicas) { diff --git a/test/framework/src/main/java/org/opensearch/test/gateway/TestShardBatchGatewayAllocator.java b/test/framework/src/main/java/org/opensearch/test/gateway/TestShardBatchGatewayAllocator.java index 156b1d7c620e6..c2ff228a6bf3a 100644 --- a/test/framework/src/main/java/org/opensearch/test/gateway/TestShardBatchGatewayAllocator.java +++ b/test/framework/src/main/java/org/opensearch/test/gateway/TestShardBatchGatewayAllocator.java @@ -10,6 +10,7 @@ import org.opensearch.cluster.node.DiscoveryNode; import org.opensearch.cluster.node.DiscoveryNodes; +import org.opensearch.cluster.routing.RerouteService; import org.opensearch.cluster.routing.ShardRouting; import org.opensearch.cluster.routing.allocation.AllocateUnassignedDecision; import org.opensearch.cluster.routing.allocation.RoutingAllocation; @@ -39,12 +40,13 @@ public TestShardBatchGatewayAllocator() { } - public TestShardBatchGatewayAllocator(CountDownLatch latch) { + public TestShardBatchGatewayAllocator(CountDownLatch latch, long maxBatchSize, RerouteService rerouteService) { + super(maxBatchSize, rerouteService); this.latch = latch; } public TestShardBatchGatewayAllocator(long maxBatchSize) { - super(maxBatchSize); + super(maxBatchSize, null); } Map<String /* node id */, Map<ShardId, ShardRouting>> knownAllocations = new HashMap<>(); From e853216ab22e50057867e1c0a5ae8bd67a6daf6c Mon Sep 17 00:00:00 2001 From: Andriy Redko <andriy.redko@aiven.io> Date: Thu, 5 Sep 2024 16:43:01 -0400 Subject: [PATCH 4/7] Add 2.17 release notes (#15762) (#15768) Signed-off-by: Andriy Redko <andriy.redko@aiven.io> --- CHANGELOG.md | 90 +-------------- .../opensearch.release-notes-2.17.0.md | 104 ++++++++++++++++++ 2 files changed, 105 insertions(+), 89 deletions(-) create mode 100644 release-notes/opensearch.release-notes-2.17.0.md diff --git a/CHANGELOG.md b/CHANGELOG.md index 4dbc667fcdbbf..435df7bda62a9 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -5,108 +5,20 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), ## [Unreleased 2.x] ### Added -- [Workload Management] Add Settings for Workload Management feature ([#15028](https://github.com/opensearch-project/OpenSearch/pull/15028)) -- Fix for hasInitiatedFetching to fix allocation explain and manual reroute APIs (([#14972](https://github.com/opensearch-project/OpenSearch/pull/14972)) -- [Workload Management] Add queryGroupId to Task ([14708](https://github.com/opensearch-project/OpenSearch/pull/14708)) -- Add setting to ignore throttling nodes for allocation of unassigned primaries in remote restore ([#14991](https://github.com/opensearch-project/OpenSearch/pull/14991)) -- [Workload Management] Add Delete QueryGroup API Logic ([#14735](https://github.com/opensearch-project/OpenSearch/pull/14735)) -- [Streaming Indexing] Enhance RestClient with a new streaming API support ([#14437](https://github.com/opensearch-project/OpenSearch/pull/14437)) -- Add basic aggregation support for derived fields ([#14618](https://github.com/opensearch-project/OpenSearch/pull/14618)) -- [Workload Management] Add Create QueryGroup API Logic ([#14680](https://github.com/opensearch-project/OpenSearch/pull/14680))- [Workload Management] Add Create QueryGroup API Logic ([#14680](https://github.com/opensearch-project/OpenSearch/pull/14680)) -- Add ThreadContextPermission for markAsSystemContext and allow core to perform the method ([#15016](https://github.com/opensearch-project/OpenSearch/pull/15016)) -- Add ThreadContextPermission for stashAndMergeHeaders and stashWithOrigin ([#15039](https://github.com/opensearch-project/OpenSearch/pull/15039)) -- [Concurrent Segment Search] Support composite aggregations with scripting ([#15072](https://github.com/opensearch-project/OpenSearch/pull/15072)) -- Add `rangeQuery` and `regexpQuery` for `constant_keyword` field type ([#14711](https://github.com/opensearch-project/OpenSearch/pull/14711)) -- Add took time to request nodes stats ([#15054](https://github.com/opensearch-project/OpenSearch/pull/15054)) -- [Workload Management] Add Get QueryGroup API Logic ([14709](https://github.com/opensearch-project/OpenSearch/pull/14709)) -- [Workload Management] Add Update QueryGroup API Logic ([#14775](https://github.com/opensearch-project/OpenSearch/pull/14775)) -- [Workload Management] QueryGroup resource tracking framework changes ([#13897](https://github.com/opensearch-project/OpenSearch/pull/13897)) -- Support filtering on a large list encoded by bitmap ([#14774](https://github.com/opensearch-project/OpenSearch/pull/14774)) -- Add slice execution listeners to SearchOperationListener interface ([#15153](https://github.com/opensearch-project/OpenSearch/pull/15153)) -- Make balanced shards allocator timebound ([#15239](https://github.com/opensearch-project/OpenSearch/pull/15239)) -- Add allowlist setting for ingest-geoip and ingest-useragent ([#15325](https://github.com/opensearch-project/OpenSearch/pull/15325)) -- Adding access to noSubMatches and noOverlappingMatches in Hyphenation ([#13895](https://github.com/opensearch-project/OpenSearch/pull/13895)) -- Star tree mapping changes ([#14605](https://github.com/opensearch-project/OpenSearch/pull/14605)) -- Add support for index level max slice count setting for concurrent segment search ([#15336](https://github.com/opensearch-project/OpenSearch/pull/15336)) -- Support cancellation for cat shards and node stats API.([#13966](https://github.com/opensearch-project/OpenSearch/pull/13966)) -- [Streaming Indexing] Introduce bulk HTTP API streaming flavor ([#15381](https://github.com/opensearch-project/OpenSearch/pull/15381)) -- Add support for centralize snapshot creation with pinned timestamp ([#15124](https://github.com/opensearch-project/OpenSearch/pull/15124)) -- Add concurrent search support for Derived Fields ([#15326](https://github.com/opensearch-project/OpenSearch/pull/15326)) -- [Workload Management] Add query group stats constructs ([#15343](https://github.com/opensearch-project/OpenSearch/pull/15343))) -- Add limit on number of processors for Ingest pipeline([#15460](https://github.com/opensearch-project/OpenSearch/pull/15465)). -- Add runAs to Subject interface and introduce IdentityAwarePlugin extension point ([#14630](https://github.com/opensearch-project/OpenSearch/pull/14630)) -- [Workload Management] Add rejection logic for co-ordinator and shard level requests ([#15428](https://github.com/opensearch-project/OpenSearch/pull/15428))) -- Adding translog durability validation in index templates ([#15494](https://github.com/opensearch-project/OpenSearch/pull/15494)) -- [Range Queries] Add new approximateable query framework to short-circuit range queries ([#13788](https://github.com/opensearch-project/OpenSearch/pull/13788)) -- [Workload Management] Add query group level failure tracking ([#15227](https://github.com/opensearch-project/OpenSearch/pull/15527)) -- [Reader Writer Separation] Add experimental search replica shard type to achieve reader writer separation ([#15237](https://github.com/opensearch-project/OpenSearch/pull/15237)) -- Add index creation using the context field ([#15290](https://github.com/opensearch-project/OpenSearch/pull/15290)) -- [Remote Publication] Add remote download stats ([#15291](https://github.com/opensearch-project/OpenSearch/pull/15291))) -- Add support to upload snapshot shard blobs with hashed prefix ([#15426](https://github.com/opensearch-project/OpenSearch/pull/15426)) -- Add prefix support to hashed prefix & infix path types on remote store ([#15557](https://github.com/opensearch-project/OpenSearch/pull/15557)) -- Add canRemain method to TargetPoolAllocationDecider to move shards from local to remote pool for hot to warm tiering ([#15010](https://github.com/opensearch-project/OpenSearch/pull/15010)) -- Add support for pluggable deciders for concurrent search ([#15363](https://github.com/opensearch-project/OpenSearch/pull/15363)) -- Add support for comma-separated list of index names to be used with Snapshot Status API ([#15409](https://github.com/opensearch-project/OpenSearch/pull/15409))[SnapshotV2] Snapshot Status API changes (#15409)) -- Optimise snapshot deletion to speed up snapshot deletion and creation ([#15568](https://github.com/opensearch-project/OpenSearch/pull/15568)) -- [Remote Publication] Added checksum validation for cluster state behind a cluster setting ([#15218](https://github.com/opensearch-project/OpenSearch/pull/15218)) -- Relax the join validation for Remote State publication ([#15471](https://github.com/opensearch-project/OpenSearch/pull/15471)) -- Optimize NodeIndicesStats output behind flag ([#14454](https://github.com/opensearch-project/OpenSearch/pull/14454)) - MultiTermQueries in keyword fields now default to `indexed` approach and gated behind cluster setting ([#15637](https://github.com/opensearch-project/OpenSearch/pull/15637)) ### Dependencies -- Bump `netty` from 4.1.111.Final to 4.1.112.Final ([#15081](https://github.com/opensearch-project/OpenSearch/pull/15081)) -- Bump `org.apache.commons:commons-lang3` from 3.14.0 to 3.15.0 ([#14861](https://github.com/opensearch-project/OpenSearch/pull/14861)) -- OpenJDK Update (July 2024 Patch releases) ([#14998](https://github.com/opensearch-project/OpenSearch/pull/14998)) -- Bump `com.microsoft.azure:msal4j` from 1.16.1 to 1.17.0 ([#14995](https://github.com/opensearch-project/OpenSearch/pull/14995), [#15420](https://github.com/opensearch-project/OpenSearch/pull/15420)) -- Bump `actions/github-script` from 6 to 7 ([#14997](https://github.com/opensearch-project/OpenSearch/pull/14997)) -- Bump `org.tukaani:xz` from 1.9 to 1.10 ([#15110](https://github.com/opensearch-project/OpenSearch/pull/15110)) -- Bump `org.apache.avro:avro` from 1.11.3 to 1.12.0 in /plugins/repository-hdfs ([#15119](https://github.com/opensearch-project/OpenSearch/pull/15119)) -- Bump `org.bouncycastle:bcpg-fips` from 1.0.7.1 to 2.0.9 ([#15103](https://github.com/opensearch-project/OpenSearch/pull/15103), [#15299](https://github.com/opensearch-project/OpenSearch/pull/15299)) -- Bump `com.azure:azure-core` from 1.49.1 to 1.51.0 ([#15111](https://github.com/opensearch-project/OpenSearch/pull/15111)) -- Bump `org.xerial.snappy:snappy-java` from 1.1.10.5 to 1.1.10.6 ([#15207](https://github.com/opensearch-project/OpenSearch/pull/15207)) -- Bump `com.azure:azure-xml` from 1.0.0 to 1.1.0 ([#15206](https://github.com/opensearch-project/OpenSearch/pull/15206)) -- Bump `reactor` from 3.5.19 to 3.5.20 ([#15262](https://github.com/opensearch-project/OpenSearch/pull/15262)) -- Bump `reactor-netty` from 1.1.21 to 1.1.22 ([#15262](https://github.com/opensearch-project/OpenSearch/pull/15262)) -- Bump `org.apache.kerby:kerb-admin` from 2.0.3 to 2.1.0 ([#15301](https://github.com/opensearch-project/OpenSearch/pull/15301)) -- Bump `com.azure:azure-core-http-netty` from 1.15.1 to 1.15.3 ([#15300](https://github.com/opensearch-project/OpenSearch/pull/15300)) -- Bump `com.gradle.develocity` from 3.17.6 to 3.18 ([#15297](https://github.com/opensearch-project/OpenSearch/pull/15297)) -- Bump `commons-cli:commons-cli` from 1.8.0 to 1.9.0 ([#15298](https://github.com/opensearch-project/OpenSearch/pull/15298)) -- Bump `opentelemetry` from 1.40.0 to 1.41.0 ([#15361](https://github.com/opensearch-project/OpenSearch/pull/15361)) -- Bump `opentelemetry-semconv` from 1.26.0-alpha to 1.27.0-alpha ([#15361](https://github.com/opensearch-project/OpenSearch/pull/15361)) -- Bump `tj-actions/changed-files` from 44 to 45 ([#15422](https://github.com/opensearch-project/OpenSearch/pull/15422)) -- Bump `dnsjava:dnsjava` from 3.6.0 to 3.6.1 ([#15418](https://github.com/opensearch-project/OpenSearch/pull/15418)) -- Bump `com.netflix.nebula.ospackage-base` from 11.9.1 to 11.10.0 ([#15419](https://github.com/opensearch-project/OpenSearch/pull/15419)) -- Bump `org.roaringbitmap:RoaringBitmap` from 1.1.0 to 1.2.1 ([#15423](https://github.com/opensearch-project/OpenSearch/pull/15423)) -- Bump `icu4j` from 70.1 to 75.1 ([#15469](https://github.com/opensearch-project/OpenSearch/pull/15469)) - Bump `com.azure:azure-identity` from 1.13.0 to 1.13.2 ([#15578](https://github.com/opensearch-project/OpenSearch/pull/15578)) ### Changed -- Add lower limit for primary and replica batch allocators timeout ([#14979](https://github.com/opensearch-project/OpenSearch/pull/14979)) -- Optimize regexp-based include/exclude on aggregations when pattern matches prefixes ([#14371](https://github.com/opensearch-project/OpenSearch/pull/14371)) -- Replace and block usages of org.apache.logging.log4j.util.Strings ([#15238](https://github.com/opensearch-project/OpenSearch/pull/15238)) -- Remote publication using minimum node version for backward compatibility ([#15216](https://github.com/opensearch-project/OpenSearch/pull/15216)) ### Deprecated ### Removed -- Remove some unused code in the search backpressure package ([#15518](https://github.com/opensearch-project/OpenSearch/pull/15518)) ### Fixed -- Fix constraint bug which allows more primary shards than average primary shards per index ([#14908](https://github.com/opensearch-project/OpenSearch/pull/14908)) -- Fix NPE when bulk ingest with empty pipeline ([#15033](https://github.com/opensearch-project/OpenSearch/pull/15033)) -- Fix missing value of FieldSort for unsigned_long ([#14963](https://github.com/opensearch-project/OpenSearch/pull/14963)) -- Fix delete index template failed when the index template matches a data stream but is unused ([#15080](https://github.com/opensearch-project/OpenSearch/pull/15080)) -- Fix array_index_out_of_bounds_exception when indexing documents with field name containing only dot ([#15126](https://github.com/opensearch-project/OpenSearch/pull/15126)) -- Fixed array field name omission in flat_object function for nested JSON ([#13620](https://github.com/opensearch-project/OpenSearch/pull/13620)) -- Fix incorrect parameter names in MinHash token filter configuration handling ([#15233](https://github.com/opensearch-project/OpenSearch/pull/15233)) -- Fix range aggregation optimization ignoring top level queries ([#15287](https://github.com/opensearch-project/OpenSearch/pull/15287)) -- Fix indexing error when flat_object field is explicitly null ([#15375](https://github.com/opensearch-project/OpenSearch/pull/15375)) -- Fix split response processor not included in allowlist ([#15393](https://github.com/opensearch-project/OpenSearch/pull/15393)) -- Fix unchecked cast in dynamic action map getter ([#15394](https://github.com/opensearch-project/OpenSearch/pull/15394)) -- Fix null values indexed as "null" strings in flat_object field ([#14069](https://github.com/opensearch-project/OpenSearch/pull/14069)) -- Fix terms query on wildcard field returns nothing ([#15607](https://github.com/opensearch-project/OpenSearch/pull/15607)) ### Security -[Unreleased 2.x]: https://github.com/opensearch-project/OpenSearch/compare/2.16...2.x +[Unreleased 2.x]: https://github.com/opensearch-project/OpenSearch/compare/2.17...2.x diff --git a/release-notes/opensearch.release-notes-2.17.0.md b/release-notes/opensearch.release-notes-2.17.0.md new file mode 100644 index 0000000000000..18b7c9ac7cd68 --- /dev/null +++ b/release-notes/opensearch.release-notes-2.17.0.md @@ -0,0 +1,104 @@ +## 2024-09-17 Version 2.17.0 Release Notes + +## [2.17.0] +### Added +- [Workload Management] Add Settings for Workload Management feature ([#15028](https://github.com/opensearch-project/OpenSearch/pull/15028)) +- Fix for hasInitiatedFetching to fix allocation explain and manual reroute APIs (([#14972](https://github.com/opensearch-project/OpenSearch/pull/14972)) +- [Workload Management] Add queryGroupId to Task ([14708](https://github.com/opensearch-project/OpenSearch/pull/14708)) +- Add setting to ignore throttling nodes for allocation of unassigned primaries in remote restore ([#14991](https://github.com/opensearch-project/OpenSearch/pull/14991)) +- [Workload Management] Add Delete QueryGroup API Logic ([#14735](https://github.com/opensearch-project/OpenSearch/pull/14735)) +- [Streaming Indexing] Enhance RestClient with a new streaming API support ([#14437](https://github.com/opensearch-project/OpenSearch/pull/14437)) +- Add basic aggregation support for derived fields ([#14618](https://github.com/opensearch-project/OpenSearch/pull/14618)) +- [Workload Management] Add Create QueryGroup API Logic ([#14680](https://github.com/opensearch-project/OpenSearch/pull/14680))- [Workload Management] Add Create QueryGroup API Logic ([#14680](https://github.com/opensearch-project/OpenSearch/pull/14680)) +- Add ThreadContextPermission for markAsSystemContext and allow core to perform the method ([#15016](https://github.com/opensearch-project/OpenSearch/pull/15016)) +- Add ThreadContextPermission for stashAndMergeHeaders and stashWithOrigin ([#15039](https://github.com/opensearch-project/OpenSearch/pull/15039)) +- [Concurrent Segment Search] Support composite aggregations with scripting ([#15072](https://github.com/opensearch-project/OpenSearch/pull/15072)) +- Add `rangeQuery` and `regexpQuery` for `constant_keyword` field type ([#14711](https://github.com/opensearch-project/OpenSearch/pull/14711)) +- Add took time to request nodes stats ([#15054](https://github.com/opensearch-project/OpenSearch/pull/15054)) +- [Workload Management] Add Get QueryGroup API Logic ([14709](https://github.com/opensearch-project/OpenSearch/pull/14709)) +- [Workload Management] Add Update QueryGroup API Logic ([#14775](https://github.com/opensearch-project/OpenSearch/pull/14775)) +- [Workload Management] QueryGroup resource tracking framework changes ([#13897](https://github.com/opensearch-project/OpenSearch/pull/13897)) +- Support filtering on a large list encoded by bitmap ([#14774](https://github.com/opensearch-project/OpenSearch/pull/14774)) +- Add slice execution listeners to SearchOperationListener interface ([#15153](https://github.com/opensearch-project/OpenSearch/pull/15153)) +- Make balanced shards allocator timebound ([#15239](https://github.com/opensearch-project/OpenSearch/pull/15239)) +- Add allowlist setting for ingest-geoip and ingest-useragent ([#15325](https://github.com/opensearch-project/OpenSearch/pull/15325)) +- Adding access to noSubMatches and noOverlappingMatches in Hyphenation ([#13895](https://github.com/opensearch-project/OpenSearch/pull/13895)) +- Star tree mapping changes ([#14605](https://github.com/opensearch-project/OpenSearch/pull/14605)) +- Add support for index level max slice count setting for concurrent segment search ([#15336](https://github.com/opensearch-project/OpenSearch/pull/15336)) +- Support cancellation for cat shards and node stats API.([#13966](https://github.com/opensearch-project/OpenSearch/pull/13966)) +- [Streaming Indexing] Introduce bulk HTTP API streaming flavor ([#15381](https://github.com/opensearch-project/OpenSearch/pull/15381)) +- Add support for centralize snapshot creation with pinned timestamp ([#15124](https://github.com/opensearch-project/OpenSearch/pull/15124)) +- Add concurrent search support for Derived Fields ([#15326](https://github.com/opensearch-project/OpenSearch/pull/15326)) +- [Workload Management] Add query group stats constructs ([#15343](https://github.com/opensearch-project/OpenSearch/pull/15343))) +- Add limit on number of processors for Ingest pipeline([#15460](https://github.com/opensearch-project/OpenSearch/pull/15465)). +- Add runAs to Subject interface and introduce IdentityAwarePlugin extension point ([#14630](https://github.com/opensearch-project/OpenSearch/pull/14630)) +- [Workload Management] Add rejection logic for co-ordinator and shard level requests ([#15428](https://github.com/opensearch-project/OpenSearch/pull/15428))) +- Adding translog durability validation in index templates ([#15494](https://github.com/opensearch-project/OpenSearch/pull/15494)) +- [Range Queries] Add new approximateable query framework to short-circuit range queries ([#13788](https://github.com/opensearch-project/OpenSearch/pull/13788)) +- [Workload Management] Add query group level failure tracking ([#15227](https://github.com/opensearch-project/OpenSearch/pull/15527)) +- [Reader Writer Separation] Add experimental search replica shard type to achieve reader writer separation ([#15237](https://github.com/opensearch-project/OpenSearch/pull/15237)) +- Add index creation using the context field ([#15290](https://github.com/opensearch-project/OpenSearch/pull/15290)) +- [Remote Publication] Add remote download stats ([#15291](https://github.com/opensearch-project/OpenSearch/pull/15291)) +- Add support to upload snapshot shard blobs with hashed prefix ([#15426](https://github.com/opensearch-project/OpenSearch/pull/15426)) +- Add prefix support to hashed prefix & infix path types on remote store ([#15557](https://github.com/opensearch-project/OpenSearch/pull/15557)) +- Add canRemain method to TargetPoolAllocationDecider to move shards from local to remote pool for hot to warm tiering ([#15010](https://github.com/opensearch-project/OpenSearch/pull/15010)) +- Add support for pluggable deciders for concurrent search ([#15363](https://github.com/opensearch-project/OpenSearch/pull/15363)) +- Optimise snapshot deletion to speed up snapshot deletion and creation ([#15568](https://github.com/opensearch-project/OpenSearch/pull/15568)) +- [Remote Publication] Added checksum validation for cluster state behind a cluster setting ([#15218](https://github.com/opensearch-project/OpenSearch/pull/15218)) +- Optimize NodeIndicesStats output behind flag ([#14454](https://github.com/opensearch-project/OpenSearch/pull/14454)) +- Add support for comma-separated list of index names to be used with Snapshot Status API ([#15409](https://github.com/opensearch-project/OpenSearch/pull/15409))[SnapshotV2] Snapshot Status API changes (#15409)) +- ClusterManagerTaskThrottler Improvements ([#15508](https://github.com/opensearch-project/OpenSearch/pull/15508)) +- Relax the join validation for Remote State publication ([#15471](https://github.com/opensearch-project/OpenSearch/pull/15471)) +- Reset DiscoveryNodes in all transport node actions request ([#15131](https://github.com/opensearch-project/OpenSearch/pull/15131)) + +### Dependencies +- Bump `netty` from 4.1.111.Final to 4.1.112.Final ([#15081](https://github.com/opensearch-project/OpenSearch/pull/15081)) +- Bump `org.apache.commons:commons-lang3` from 3.14.0 to 3.15.0 ([#14861](https://github.com/opensearch-project/OpenSearch/pull/14861)) +- OpenJDK Update (July 2024 Patch releases) ([#14998](https://github.com/opensearch-project/OpenSearch/pull/14998)) +- Bump `com.microsoft.azure:msal4j` from 1.16.1 to 1.17.0 ([#14995](https://github.com/opensearch-project/OpenSearch/pull/14995), [#15420](https://github.com/opensearch-project/OpenSearch/pull/15420)) +- Bump `actions/github-script` from 6 to 7 ([#14997](https://github.com/opensearch-project/OpenSearch/pull/14997)) +- Bump `org.tukaani:xz` from 1.9 to 1.10 ([#15110](https://github.com/opensearch-project/OpenSearch/pull/15110)) +- Bump `org.apache.avro:avro` from 1.11.3 to 1.12.0 in /plugins/repository-hdfs ([#15119](https://github.com/opensearch-project/OpenSearch/pull/15119)) +- Bump `org.bouncycastle:bcpg-fips` from 1.0.7.1 to 2.0.9 ([#15103](https://github.com/opensearch-project/OpenSearch/pull/15103), [#15299](https://github.com/opensearch-project/OpenSearch/pull/15299)) +- Bump `com.azure:azure-core` from 1.49.1 to 1.51.0 ([#15111](https://github.com/opensearch-project/OpenSearch/pull/15111)) +- Bump `org.xerial.snappy:snappy-java` from 1.1.10.5 to 1.1.10.6 ([#15207](https://github.com/opensearch-project/OpenSearch/pull/15207)) +- Bump `com.azure:azure-xml` from 1.0.0 to 1.1.0 ([#15206](https://github.com/opensearch-project/OpenSearch/pull/15206)) +- Bump `reactor` from 3.5.19 to 3.5.20 ([#15262](https://github.com/opensearch-project/OpenSearch/pull/15262)) +- Bump `reactor-netty` from 1.1.21 to 1.1.22 ([#15262](https://github.com/opensearch-project/OpenSearch/pull/15262)) +- Bump `org.apache.kerby:kerb-admin` from 2.0.3 to 2.1.0 ([#15301](https://github.com/opensearch-project/OpenSearch/pull/15301)) +- Bump `com.azure:azure-core-http-netty` from 1.15.1 to 1.15.3 ([#15300](https://github.com/opensearch-project/OpenSearch/pull/15300)) +- Bump `com.gradle.develocity` from 3.17.6 to 3.18 ([#15297](https://github.com/opensearch-project/OpenSearch/pull/15297)) +- Bump `commons-cli:commons-cli` from 1.8.0 to 1.9.0 ([#15298](https://github.com/opensearch-project/OpenSearch/pull/15298)) +- Bump `opentelemetry` from 1.40.0 to 1.41.0 ([#15361](https://github.com/opensearch-project/OpenSearch/pull/15361)) +- Bump `opentelemetry-semconv` from 1.26.0-alpha to 1.27.0-alpha ([#15361](https://github.com/opensearch-project/OpenSearch/pull/15361)) +- Bump `tj-actions/changed-files` from 44 to 45 ([#15422](https://github.com/opensearch-project/OpenSearch/pull/15422)) +- Bump `dnsjava:dnsjava` from 3.6.0 to 3.6.1 ([#15418](https://github.com/opensearch-project/OpenSearch/pull/15418)) +- Bump `com.netflix.nebula.ospackage-base` from 11.9.1 to 11.10.0 ([#15419](https://github.com/opensearch-project/OpenSearch/pull/15419)) +- Bump `org.roaringbitmap:RoaringBitmap` from 1.1.0 to 1.2.1 ([#15423](https://github.com/opensearch-project/OpenSearch/pull/15423)) +- Bump `icu4j` from 70.1 to 75.1 ([#15469](https://github.com/opensearch-project/OpenSearch/pull/15469)) + +### Changed +- Add lower limit for primary and replica batch allocators timeout ([#14979](https://github.com/opensearch-project/OpenSearch/pull/14979)) +- Optimize regexp-based include/exclude on aggregations when pattern matches prefixes ([#14371](https://github.com/opensearch-project/OpenSearch/pull/14371)) +- Replace and block usages of org.apache.logging.log4j.util.Strings ([#15238](https://github.com/opensearch-project/OpenSearch/pull/15238)) +- Remote publication using minimum node version for backward compatibility ([#15216](https://github.com/opensearch-project/OpenSearch/pull/15216)) + +### Deprecated + +### Removed +- Remove some unused code in the search backpressure package ([#15518](https://github.com/opensearch-project/OpenSearch/pull/15518)) + +### Fixed +- Fix constraint bug which allows more primary shards than average primary shards per index ([#14908](https://github.com/opensearch-project/OpenSearch/pull/14908)) +- Fix NPE when bulk ingest with empty pipeline ([#15033](https://github.com/opensearch-project/OpenSearch/pull/15033)) +- Fix missing value of FieldSort for unsigned_long ([#14963](https://github.com/opensearch-project/OpenSearch/pull/14963)) +- Fix delete index template failed when the index template matches a data stream but is unused ([#15080](https://github.com/opensearch-project/OpenSearch/pull/15080)) +- Fix array_index_out_of_bounds_exception when indexing documents with field name containing only dot ([#15126](https://github.com/opensearch-project/OpenSearch/pull/15126)) +- Fixed array field name omission in flat_object function for nested JSON ([#13620](https://github.com/opensearch-project/OpenSearch/pull/13620)) +- Fix incorrect parameter names in MinHash token filter configuration handling ([#15233](https://github.com/opensearch-project/OpenSearch/pull/15233)) +- Fix range aggregation optimization ignoring top level queries ([#15287](https://github.com/opensearch-project/OpenSearch/pull/15287)) +- Fix indexing error when flat_object field is explicitly null ([#15375](https://github.com/opensearch-project/OpenSearch/pull/15375)) +- Fix split response processor not included in allowlist ([#15393](https://github.com/opensearch-project/OpenSearch/pull/15393)) +- Fix unchecked cast in dynamic action map getter ([#15394](https://github.com/opensearch-project/OpenSearch/pull/15394)) +- Fix null values indexed as "null" strings in flat_object field ([#14069](https://github.com/opensearch-project/OpenSearch/pull/14069)) +- Fix terms query on wildcard field returns nothing ([#15607](https://github.com/opensearch-project/OpenSearch/pull/15607)) From e6862cdc868b20afed6ac23696d5e9ae7df4906e Mon Sep 17 00:00:00 2001 From: "opensearch-trigger-bot[bot]" <98922864+opensearch-trigger-bot[bot]@users.noreply.github.com> Date: Thu, 5 Sep 2024 17:58:14 -0400 Subject: [PATCH 5/7] Adding release notes (#15771) (#15779) (cherry picked from commit e892f23c443ddfbfd82bf0e447636231d34ecd2f) Signed-off-by: Harsha Vamsi Kalluri <harshavamsi096@gmail.com> Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> --- release-notes/opensearch.release-notes-2.17.0.md | 1 + 1 file changed, 1 insertion(+) diff --git a/release-notes/opensearch.release-notes-2.17.0.md b/release-notes/opensearch.release-notes-2.17.0.md index 18b7c9ac7cd68..41c7a89ea634c 100644 --- a/release-notes/opensearch.release-notes-2.17.0.md +++ b/release-notes/opensearch.release-notes-2.17.0.md @@ -50,6 +50,7 @@ - ClusterManagerTaskThrottler Improvements ([#15508](https://github.com/opensearch-project/OpenSearch/pull/15508)) - Relax the join validation for Remote State publication ([#15471](https://github.com/opensearch-project/OpenSearch/pull/15471)) - Reset DiscoveryNodes in all transport node actions request ([#15131](https://github.com/opensearch-project/OpenSearch/pull/15131)) +- MultiTermQueries in keyword fields now default to `indexed` approach and gated behind cluster setting ([#15637](https://github.com/opensearch-project/OpenSearch/pull/15637)) ### Dependencies - Bump `netty` from 4.1.111.Final to 4.1.112.Final ([#15081](https://github.com/opensearch-project/OpenSearch/pull/15081)) From 93b0755bd64fb35c61aef0f5b223c30bfd9a84b2 Mon Sep 17 00:00:00 2001 From: "opensearch-trigger-bot[bot]" <98922864+opensearch-trigger-bot[bot]@users.noreply.github.com> Date: Thu, 5 Sep 2024 18:00:25 -0400 Subject: [PATCH 6/7] Fixing MacOS 13 assemble workflows (#15747) (#15773) (cherry picked from commit fe61e4f299d4b832dc232cb223ec946585177f52) Signed-off-by: Andriy Redko <andriy.redko@aiven.io> Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> --- .github/workflows/assemble.yml | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/.github/workflows/assemble.yml b/.github/workflows/assemble.yml index 294627622a136..b3838b8e5ae97 100644 --- a/.github/workflows/assemble.yml +++ b/.github/workflows/assemble.yml @@ -30,8 +30,11 @@ jobs: - name: Setup docker (missing on MacOS) id: setup_docker if: runner.os == 'macos' + continue-on-error: true run: | - exit 0; + brew install docker colima coreutils + gtimeout 15m colima start + shell: bash - name: Run Gradle (assemble) if: runner.os == 'macos' && steps.setup_docker.outcome != 'success' run: | @@ -45,4 +48,4 @@ jobs: - name: Run Gradle (assemble) if: runner.os == 'macos' && steps.setup_docker.outcome == 'success' run: | - exit 0; + ./gradlew assemble --parallel --no-build-cache -PDISABLE_BUILD_CACHE -Druntime.java=${{ matrix.java }} From 3b8a741a8547c21317449b7d783edd378a8eecb7 Mon Sep 17 00:00:00 2001 From: Harsha Vamsi Kalluri <harshavamsi096@gmail.com> Date: Thu, 5 Sep 2024 15:51:01 -0700 Subject: [PATCH 7/7] Remove un-necessary changelog entry (#15781) Signed-off-by: Harsha Vamsi Kalluri <harshavamsi096@gmail.com> --- CHANGELOG.md | 1 - 1 file changed, 1 deletion(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index 435df7bda62a9..6397b81a5f1c3 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -5,7 +5,6 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), ## [Unreleased 2.x] ### Added -- MultiTermQueries in keyword fields now default to `indexed` approach and gated behind cluster setting ([#15637](https://github.com/opensearch-project/OpenSearch/pull/15637)) ### Dependencies - Bump `com.azure:azure-identity` from 1.13.0 to 1.13.2 ([#15578](https://github.com/opensearch-project/OpenSearch/pull/15578))