[SPARK-8567] [SQL] Increase the timeout of HiveSparkSubmitSuite

[yhuai]

https://issues.apache.org/jira/browse/SPARK-8567

[SparkQA]

Test build #35567 has finished for PR 6957 at commit f8aab7f.

• This patch fails MiMa tests.
• This patch merges cleanly.
• This patch adds no public classes.

[JoshRosen]

Jenkins, retest this please.

[SparkQA]

Test build #35572 has finished for PR 6957 at commit f8aab7f.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[yhuai]

hmm... Jenkins still does not have the logs of applications launched in HiveSparkSubmitSuite.

[JoshRosen]

@yhuai, where are those logs written when you run the tests locally? We might have to add those log paths to the list that Jenkins checks when attaching build artifacts in the post-build step.

[yhuai]

They are in sql/hive/target/uni-tests.log. The log in jenkins only shows

12:47:58.306 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO HiveSparkSubmitSuite: 

===== TEST OUTPUT FOR o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8368: includes jars passed in through --jars' =====

12:49:04.038 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO HiveSparkSubmitSuite: 

===== FINISHED o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8368: includes jars passed in through --jars' =====

12:49:04.039 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO HiveSparkSubmitSuite: 

===== TEST OUTPUT FOR o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8020: set sql conf in spark conf' =====

12:49:52.939 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO HiveSparkSubmitSuite: 

===== FINISHED o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8020: set sql conf in spark conf' =====

12:49:52.941 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO HiveSparkSubmitSuite: 

===== TEST OUTPUT FOR o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8489: MissingRequirementError during reflection' =====

12:50:42.262 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO HiveSparkSubmitSuite: 

===== FINISHED o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8489: MissingRequirementError during reflection' =====

[andrewor14]

I just ran this locally and my sql/hive/target/unit-test.logs looks different. Looks like they are logged to the right file there.

===== TEST OUTPUT FOR o.a.s.sql.hive.HiveSparkSubmitSuite: 'SPARK-8368: includes jars passed in through --jars' ===== 

14:21:43.037 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite INFO SparkContext: Running Spark version 1.5.0-SNAPSHOT
14:21:43.074 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite DEBUG MutableMetricsFactory: field org.apache.hadoop.metrics2.lib.MutableRate org.apache.hadoop.security.UserGroupInformation$UgiMetrics.loginSuccess with annotation @org.apache.hadoop.metrics2.annotation.Metric(valueName=Time, about=, value=[Rate of successful kerberos logins and latency (milliseconds)], always=false, type=DEFAULT, sampleName=Ops)
14:21:43.078 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite DEBUG MutableMetricsFactory: field org.apache.hadoop.metrics2.lib.MutableRate org.apache.hadoop.security.UserGroupInformation$UgiMetrics.loginFailure with annotation @org.apache.hadoop.metrics2.annotation.Metric(valueName=Time, about=, value=[Rate of failed kerberos logins and latency (milliseconds)], always=false, type=DEFAULT, sampleName=Ops)
14:21:43.078 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite DEBUG MetricsSystemImpl: UgiMetrics, User and group related metrics
14:21:43.305 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite DEBUG KerberosName: Kerberos krb5 configuration not found, setting default realm to empty 
14:21:43.307 pool-1-thread-1-ScalaTest-running-HiveSparkSubmitSuite DEBUG Groups:  Creating new Groups object
...

[yhuai]

@andrewor14 Yeah, I also got that in my laptop.

[liancheng]

Partly reproduced this issue on my laptop. Trying my luck with #6978 by printing rather than logging stdout/stderr of the spawned process.

[andrewor14]

Ok, let's retest this please now that #6978 is merged.

[SparkQA]

Test build #35705 has finished for PR 6957 at commit f8aab7f.

• This patch fails MiMa tests.
• This patch merges cleanly.
• This patch adds no public classes.

[andrewor14]

retest this please

[andrewor14]

could just make this default to false

[SparkQA]

Test build #35714 has finished for PR 6957 at commit f8aab7f.

• This patch fails Spark unit tests.
• This patch merges cleanly.
• This patch adds no public classes.

[SparkQA]

Test build #35717 has finished for PR 6957 at commit 2292dfb.

• This patch fails Spark unit tests.
• This patch merges cleanly.
• This patch adds no public classes.

[andrewor14]

LGTM as a temporary fix. Let's merge this into master and 1.4 and see if it fixes the test failures.

[andrewor14]

I will leave the JIRA open for a day or two until after we can verify that this fixes it.

[SparkQA]

Test build #35726 has finished for PR 6957 at commit 62dff5b.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • class StreamingLinearAlgorithm(object):
  • class StreamingLogisticRegressionWithSGD(StreamingLinearAlgorithm):
  • case class PrecisionInfo(precision: Int, scale: Int)

[liancheng]

@yhuai As for why the log doesn't appear on Jenkins, my guess is that the spark-submit process wasn't started properly because of, for example, some class not found error. I'd suggest to enable SPARK_PRINT_LAUNCH_COMMAND, then login into a Jenkins worker node and try exactly the same command line there. Currently, a super long classpath (it fills two and a half screens of mine...) appear on Jenkins for this test suite. This might be the reason why this error only appear on Jenkins, since classpath can be environment sensitive.