[ https://issues.apache.org/jira/browse/GOBBLIN-1071?focusedWorklogId=398531&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-398531 ]
ASF GitHub Bot logged work on GOBBLIN-1071: ------------------------------------------- Author: ASF GitHub Bot Created on: 05/Mar/20 17:24 Start Date: 05/Mar/20 17:24 Worklog Time Spent: 10m Work Description: sv2000 commented on pull request #2909: [GOBBLIN-1071] Retry task initialization URL: https://github.com/apache/incubator-gobblin/pull/2909#discussion_r388437144 ########## File path: gobblin-cluster/src/main/java/org/apache/gobblin/cluster/SingleTask.java ########## @@ -50,49 +54,74 @@ public class SingleTask { private static final Logger _logger = LoggerFactory.getLogger(SingleTask.class); + public static final String MAX_RETRY_WAITING_FOR_INIT_KEY = "maxRetryBlockedOnTaskAttemptInit"; + public static final int DEFAULT_MAX_RETRY_WAITING_FOR_INIT = 2; - private GobblinMultiTaskAttempt _taskattempt; + private GobblinMultiTaskAttempt _taskAttempt; private String _jobId; private Path _workUnitFilePath; private Path _jobStateFilePath; private FileSystem _fs; private TaskAttemptBuilder _taskAttemptBuilder; private StateStores _stateStores; private Config _dynamicConfig; + private List<WorkUnit> _workUnits; + private JobState _jobState; + // Preventing Helix calling cancel before taskAttempt is created + // Checking if taskAttempt is empty is not enough, since canceller runs in different thread as runner, the case to + // to avoided here is taskAttempt being created and start to run after cancel has been called. + private Condition _taskAttemptBuilt; + private Lock _lock; + + /** + * Do all heavy-lifting of initialization in constructor which could be retried if failed, + * see the example in {@link GobblinHelixTask}. + */ SingleTask(String jobId, Path workUnitFilePath, Path jobStateFilePath, FileSystem fs, - TaskAttemptBuilder taskAttemptBuilder, StateStores stateStores, Config dynamicConfig) { + TaskAttemptBuilder taskAttemptBuilder, StateStores stateStores, Config dynamicConfig) throws IOException { _jobId = jobId; _workUnitFilePath = workUnitFilePath; _jobStateFilePath = jobStateFilePath; _fs = fs; _taskAttemptBuilder = taskAttemptBuilder; _stateStores = stateStores; _dynamicConfig = dynamicConfig; + + _workUnits = getWorkUnits(); Review comment: Why do we have to eagerly initialize workunits? ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking ------------------- Worklog Id: (was: 398531) Time Spent: 50m (was: 40m) > Retry Task Initialization > ------------------------- > > Key: GOBBLIN-1071 > URL: https://issues.apache.org/jira/browse/GOBBLIN-1071 > Project: Apache Gobblin > Issue Type: Improvement > Reporter: Lei Sun > Priority: Major > Time Spent: 50m > Remaining Estimate: 0h > -- This message was sent by Atlassian Jira (v8.3.4#803005)