[ https://issues.apache.org/jira/browse/FLINK-8360?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16313146#comment-16313146 ]
ASF GitHub Bot commented on FLINK-8360: --------------------------------------- Github user pnowojski commented on a diff in the pull request: https://github.com/apache/flink/pull/5239#discussion_r159863616 --- Diff: flink-contrib/flink-storm/src/test/java/org/apache/flink/storm/wrappers/BoltWrapperTest.java --- @@ -375,11 +376,13 @@ public void declareOutputFields(OutputFieldsDeclarer declarer) { when(env.getMetricGroup()).thenReturn(new UnregisteredTaskMetricsGroup()); when(env.getTaskManagerInfo()).thenReturn(new TestingTaskManagerRuntimeInfo()); + final CloseableRegistry closeableRegistry = new CloseableRegistry(); StreamTask<?, ?> mockTask = mock(StreamTask.class); when(mockTask.getCheckpointLock()).thenReturn(new Object()); when(mockTask.getConfiguration()).thenReturn(new StreamConfig(new Configuration())); when(mockTask.getEnvironment()).thenReturn(env); when(mockTask.getExecutionConfig()).thenReturn(execConfig); + when(mockTask.getCancelables()).thenReturn(closeableRegistry); --- End diff -- Mockito as always! Why not replace this ugly block with `NoOpStreamTask` or something else implementing `StreamTask`? It will save the hassle for next developer that will try to call some other method from this mock, that wasn't defined properly. > Implement task-local state recovery > ----------------------------------- > > Key: FLINK-8360 > URL: https://issues.apache.org/jira/browse/FLINK-8360 > Project: Flink > Issue Type: New Feature > Components: State Backends, Checkpointing > Reporter: Stefan Richter > Assignee: Stefan Richter > Fix For: 1.5.0 > > > This issue tracks the development of recovery from task-local state. The main > idea is to have a secondary, local copy of the checkpointed state, while > there is still a primary copy in DFS that we report to the checkpoint > coordinator. > Recovery can attempt to restore from the secondary local copy, if available, > to save network bandwidth. This requires that the assignment from tasks to > slots is as sticky is possible. > For starters, we will implement this feature for all managed keyed states and > can easily enhance it to all other state types (e.g. operator state) later. -- This message was sent by Atlassian JIRA (v6.4.14#64029)