[ https://issues.apache.org/jira/browse/IGNITE-11253?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vyacheslav Koptilin updated IGNITE-11253: ----------------------------------------- Description: * In case of eager TTL is configured, a starting node creates and starts {{cleanupWorker}} (see {{GridCacheTtlManager.start0()}}) * {{GridCacheSharedTtlCleanupManager.CleanupWorker}}, in its turn, has to wait for {{discovery().localJoin()}} future that is completed by discovery thread. * On the other hand, the exchange thread stops cache contexts and, therefore, it stops the {{cleanupWorker}} as well. {code:java} org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager.stopCleanupWorker(GridCacheSharedTtlCleanupManager.java:109) org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager.unregister(GridCacheSharedTtlCleanupManager.java:82) org.apache.ignite.internal.processors.cache.GridCacheTtlManager.onKernalStop0(GridCacheTtlManager.java:110) org.apache.ignite.internal.processors.cache.GridCacheManagerAdapter.onKernalStop(GridCacheManagerAdapter.java:111) org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStop(GridCacheProcessor.java:1495) org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStopCaches(GridCacheProcessor.java:1182) org.apache.ignite.internal.processors.cache.GridCacheProcessor$CacheRecoveryLifecycle.onBaselineChange(GridCacheProcessor.java:5637) org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionsExchangeFuture.initCachesOnLocalJoin(GridDhtPartitionsExchangeFuture.java:910) org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionsExchangeFuture.init(GridDhtPartitionsExchangeFuture.java:792) {code} So, exchange thread may try to stop the {{cleanupWorker}} before the {{localJoin}} future is completed by discovery thread. Unfortunately, `cleanupWorker` incorrectly handles this situation, and this fact can lead to a node failure: {code:java} Critical system error detected. Will be handled accordingly to configured handler [hnd=StopNodeFailureHandler [super=AbstractFailureHandler [ignoredFailureTypes=UnmodifiableSet [SYSTEM_WORKER_BLOCKED, SYSTEM_CRITICAL_OPERATION_TIMEOUT]]], failureCtx=FailureContext [type=SYSTEM_WORKER_TERMINATION, err=class o.a.i.IgniteException: Got interrupted while waiting for future to complete.]] class org.apache.ignite.IgniteException: Got interrupted while waiting for future to complete. at org.apache.ignite.internal.managers.discovery.GridDiscoveryManager.localJoin(GridDiscoveryManager.java:2217) at org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager$CleanupWorker.body(GridCacheSharedTtlCleanupManager.java:136) at org.apache.ignite.internal.util.worker.GridWorker.run(GridWorker.java:120) at java.lang.Thread.run(Thread.java:748) Caused by: class org.apache.ignite.internal.IgniteInterruptedCheckedException: Got interrupted while waiting for future to complete. at org.apache.ignite.internal.util.future.GridFutureAdapter.get0(GridFutureAdapter.java:186) at org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:141) at org.apache.ignite.internal.managers.discovery.GridDiscoveryManager.localJoin(GridDiscoveryManager.java:2214) ... 3 more {code} The obvious fix is changing the catch block {code:java} catch (Throwable t) { if (!(t instanceof IgniteInterruptedCheckedException)) err = t; throw t; } {code} to the following: {code:java} catch (Throwable t) { if (!(X.cause(t, IgniteInterruptedCheckedException.class))) err = t; throw t; } {code} was: * In case of eager TTL is configured, a starting node creates and starts {{cleanupWorker}} (see {{GridCacheTtlManager.start0()}}) * {{GridCacheSharedTtlCleanupManager.CleanupWorker}}, in its turn, has to wait for {{discovery().localJoin()}} future that is completed by discovery thread. * On the other hand, the exchange thread stops cache contexts and, therefore, it stops the \{{cleanupWorker}} as well. {code:java} org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager.stopCleanupWorker(GridCacheSharedTtlCleanupManager.java:109) org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager.unregister(GridCacheSharedTtlCleanupManager.java:82) org.apache.ignite.internal.processors.cache.GridCacheTtlManager.onKernalStop0(GridCacheTtlManager.java:110) org.apache.ignite.internal.processors.cache.GridCacheManagerAdapter.onKernalStop(GridCacheManagerAdapter.java:111) org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStop(GridCacheProcessor.java:1495) org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStopCaches(GridCacheProcessor.java:1182) org.apache.ignite.internal.processors.cache.GridCacheProcessor$CacheRecoveryLifecycle.onBaselineChange(GridCacheProcessor.java:5637) org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionsExchangeFuture.initCachesOnLocalJoin(GridDhtPartitionsExchangeFuture.java:910) org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionsExchangeFuture.init(GridDhtPartitionsExchangeFuture.java:792) {code} So, exchange thread may try to stop the {{cleanupWorker}} before the {{localJoin}} future is completed by discovery thread. Unfortunately, `cleanupWorker` incorrectly handles this situation, and this fact can lead to a node failure: {code:java} Critical system error detected. Will be handled accordingly to configured handler [hnd=StopNodeFailureHandler [super=AbstractFailureHandler [ignoredFailureTypes=UnmodifiableSet [SYSTEM_WORKER_BLOCKED, SYSTEM_CRITICAL_OPERATION_TIMEOUT]]], failureCtx=FailureContext [type=SYSTEM_WORKER_TERMINATION, err=class o.a.i.IgniteException: Got interrupted while waiting for future to complete.]] class org.apache.ignite.IgniteException: Got interrupted while waiting for future to complete. at org.apache.ignite.internal.managers.discovery.GridDiscoveryManager.localJoin(GridDiscoveryManager.java:2217) at org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager$CleanupWorker.body(GridCacheSharedTtlCleanupManager.java:136) at org.apache.ignite.internal.util.worker.GridWorker.run(GridWorker.java:120) at java.lang.Thread.run(Thread.java:748) Caused by: class org.apache.ignite.internal.IgniteInterruptedCheckedException: Got interrupted while waiting for future to complete. at org.apache.ignite.internal.util.future.GridFutureAdapter.get0(GridFutureAdapter.java:186) at org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:141) at org.apache.ignite.internal.managers.discovery.GridDiscoveryManager.localJoin(GridDiscoveryManager.java:2214) ... 3 more {code} > When a node that is not part of the base topology joins the cluster, it may > lead to a node failure. > --------------------------------------------------------------------------------------------------- > > Key: IGNITE-11253 > URL: https://issues.apache.org/jira/browse/IGNITE-11253 > Project: Ignite > Issue Type: Bug > Affects Versions: 2.7 > Reporter: Vyacheslav Koptilin > Assignee: Vyacheslav Koptilin > Priority: Major > Fix For: 2.8 > > > * In case of eager TTL is configured, a starting node creates and starts > {{cleanupWorker}} (see {{GridCacheTtlManager.start0()}}) > * {{GridCacheSharedTtlCleanupManager.CleanupWorker}}, in its turn, has to > wait for {{discovery().localJoin()}} future that is completed by discovery > thread. > * On the other hand, the exchange thread stops cache contexts and, > therefore, it stops the {{cleanupWorker}} as well. > > {code:java} > org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager.stopCleanupWorker(GridCacheSharedTtlCleanupManager.java:109) > org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager.unregister(GridCacheSharedTtlCleanupManager.java:82) > org.apache.ignite.internal.processors.cache.GridCacheTtlManager.onKernalStop0(GridCacheTtlManager.java:110) > org.apache.ignite.internal.processors.cache.GridCacheManagerAdapter.onKernalStop(GridCacheManagerAdapter.java:111) > org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStop(GridCacheProcessor.java:1495) > org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStopCaches(GridCacheProcessor.java:1182) > org.apache.ignite.internal.processors.cache.GridCacheProcessor$CacheRecoveryLifecycle.onBaselineChange(GridCacheProcessor.java:5637) > org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionsExchangeFuture.initCachesOnLocalJoin(GridDhtPartitionsExchangeFuture.java:910) > org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionsExchangeFuture.init(GridDhtPartitionsExchangeFuture.java:792) > {code} > So, exchange thread may try to stop the {{cleanupWorker}} before the > {{localJoin}} future is completed by discovery thread. Unfortunately, > `cleanupWorker` incorrectly handles this situation, and this fact can lead to > a node failure: > {code:java} > Critical system error detected. Will be handled accordingly to configured > handler [hnd=StopNodeFailureHandler [super=AbstractFailureHandler > [ignoredFailureTypes=UnmodifiableSet [SYSTEM_WORKER_BLOCKED, > SYSTEM_CRITICAL_OPERATION_TIMEOUT]]], failureCtx=FailureContext > [type=SYSTEM_WORKER_TERMINATION, err=class o.a.i.IgniteException: Got > interrupted while waiting for future to complete.]] > class org.apache.ignite.IgniteException: Got interrupted while waiting for > future to complete. > at > org.apache.ignite.internal.managers.discovery.GridDiscoveryManager.localJoin(GridDiscoveryManager.java:2217) > at > org.apache.ignite.internal.processors.cache.GridCacheSharedTtlCleanupManager$CleanupWorker.body(GridCacheSharedTtlCleanupManager.java:136) > at org.apache.ignite.internal.util.worker.GridWorker.run(GridWorker.java:120) > at java.lang.Thread.run(Thread.java:748) > Caused by: class > org.apache.ignite.internal.IgniteInterruptedCheckedException: Got interrupted > while waiting for future to complete. > at > org.apache.ignite.internal.util.future.GridFutureAdapter.get0(GridFutureAdapter.java:186) > at > org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:141) > at > org.apache.ignite.internal.managers.discovery.GridDiscoveryManager.localJoin(GridDiscoveryManager.java:2214) > ... 3 more > {code} > The obvious fix is changing the catch block > {code:java} > catch (Throwable t) { > if (!(t instanceof IgniteInterruptedCheckedException)) > err = t; > throw t; > } > {code} > to the following: > {code:java} > catch (Throwable t) { > if (!(X.cause(t, IgniteInterruptedCheckedException.class))) > err = t; > throw t; > } > {code} -- This message was sent by Atlassian JIRA (v7.6.3#76005)