You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Aihua Xu (JIRA)" <ji...@apache.org> on 2017/08/22 20:53:01 UTC
[jira] [Commented] (HIVE-8789) UT: fix udf_context_aware
[ https://issues.apache.org/jira/browse/HIVE-8789?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16137389#comment-16137389 ]
Aihua Xu commented on HIVE-8789:
--------------------------------
This test is not applicable to HoS since the counters can only be read from the driver. So such usage is not supported.
Sees http://spark.apache.org/docs/latest/rdd-programming-guide.html
> UT: fix udf_context_aware
> --------------------------
>
> Key: HIVE-8789
> URL: https://issues.apache.org/jira/browse/HIVE-8789
> Project: Hive
> Issue Type: Sub-task
> Components: Tests
> Affects Versions: spark-branch
> Reporter: Thomas Friedrich
> Assignee: Aihua Xu
> Priority: Minor
>
> The test udf_context_aware fails with a NPE here:
> Thread [main] (Suspended (exception NullPointerException))
> DummyContextUDF.evaluate(GenericUDF$DeferredObject[]) line: 42
> ExprNodeGenericFuncEvaluator._evaluate(Object, int) line: 169
> ExprNodeGenericFuncEvaluator(ExprNodeEvaluator<T>).evaluate(Object, int) line: 77
> ExprNodeGenericFuncEvaluator(ExprNodeEvaluator<T>).evaluate(Object) line: 65
> SelectOperator.processOp(Object, int) line: 77
> TableScanOperator(Operator<T>).forward(Object, ObjectInspector) line: 815
> TableScanOperator.processOp(Object, int) line: 95
> FetchOperator.pushRow(InspectableObject) line: 577
> FetchOperator.pushRow() line: 569
> FetchTask.fetch(List) line: 138
> Driver.getResults(List) line: 1661
> CliDriver.processLocalCmd(String, CommandProcessor, CliSessionState) line: 267
> CliDriver.processCmd(String) line: 199
> CliDriver.processLine(String, boolean) line: 410
> CliDriver.processLine(String) line: 345
> QTestUtil.executeClient(String) line: 832
> TestSparkCliDriver.runTest(String, String, String) line: 136
> TestSparkCliDriver.testCliDriver_udf_context_aware() line: 120
> NativeMethodAccessorImpl.invoke0(Method, Object, Object[]) line: not available [native method]
> NativeMethodAccessorImpl.invoke(Object, Object[]) line: 57
> DelegatingMethodAccessorImpl.invoke(Object, Object[]) line: 43
> Method.invoke(Object, Object...) line: 606
> TestSparkCliDriver(TestCase).runTest() line: 176
> TestSparkCliDriver(TestCase).runBare() line: 141
> TestResult$1.protect() line: 122
> TestResult.runProtected(Test, Protectable) line: 142
> TestResult.run(TestCase) line: 125
> TestSparkCliDriver(TestCase).run(TestResult) line: 129
> TestSuite.runTest(Test, TestResult) line: 255
> TestSuite.run(TestResult) line: 250
> SuiteMethod(JUnit38ClassRunner).run(RunNotifier) line: 84
> JUnit4Provider.execute(Class<?>, RunNotifier, String[]) line: 264
> JUnit4Provider.executeTestSet(Class<?>, RunListener, RunNotifier) line: 153
> JUnit4Provider.invoke(Object) line: 124
> ForkedBooter.invokeProviderInSameClassLoader(Object, Object, ProviderConfiguration, boolean, StartupConfiguration, boolean) line: 200
> ForkedBooter.runSuitesInProcess(Object, StartupConfiguration, ProviderConfiguration, PrintStream) line: 153
> ForkedBooter.main(String[]) line: 103
> While debugging I found that the MapredContext object is null here:
> Reporter reporter = context.getReporter();
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)