[SPARK-23489][SQL][TEST] HiveExternalCatalogVersionsSuite should verify the downloaded file #21210

dongjoon-hyun · 2018-05-02T05:16:04Z

What changes were proposed in this pull request?

Although SPARK-22654 made HiveExternalCatalogVersionsSuite download from Apache mirrors three times, it has been flaky because it didn't verify the downloaded file. Some Apache mirrors terminate the downloading abnormally, the corrupted file shows the following errors.

gzip: stdin: not in gzip format
tar: Child returned status 1
tar: Error is not recoverable: exiting now
22:46:32.700 WARN org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite: 

===== POSSIBLE THREAD LEAK IN SUITE o.a.s.sql.hive.HiveExternalCatalogVersionsSuite, thread names: Keep-Alive-Timer =====

*** RUN ABORTED ***
  java.io.IOException: Cannot run program "./bin/spark-submit" (in directory "/tmp/test-spark/spark-2.2.0"): error=2, No such file or directory

This has been reported weirdly in two ways. For example, the above case is reported as Case 2 no failures.

Case 1. Test Result (1 failure / +1)
Case 2. Test Result (no failures)

This PR aims to make HiveExternalCatalogVersionsSuite more robust by verifying the downloaded tgz file by extracting and checking the existence of bin/spark-submit. If it turns out that the file is empty or corrupted, HiveExternalCatalogVersionsSuite will do retry logic like the download failure.

How was this patch tested?

Pass the Jenkins.

…fy the downloaded file

gatorsmile · 2018-05-02T05:29:12Z

Thanks for working on it!

SparkQA · 2018-05-02T07:05:01Z

Test build #90029 has finished for PR 21210 at commit 51d4c0e.

This patch fails due to an unknown error code, -9.
This patch merges cleanly.
This patch adds no public classes.

hvanhovell · 2018-05-02T09:31:56Z

retest this please

SparkQA · 2018-05-02T12:05:44Z

Test build #90039 has finished for PR 21210 at commit 51d4c0e.

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

dongjoon-hyun · 2018-05-02T17:08:40Z

Thank you for review, @gatorsmile and @hvanhovell !
And, the test failure is irrelevant to this one.

dongjoon-hyun · 2018-05-02T17:08:53Z

Retest this please.

dongjoon-hyun · 2018-05-02T17:53:25Z

I'm not sure we can avoid SparkR flakiness here. Can we ignore SparkR failure like #21175 , @gatorsmile , @hvanhovell , @cloud-fan ?

In fact, the test build 90039, https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/90039/console, shows no other error except SparkR one.

SparkQA · 2018-05-02T18:39:53Z

Test build #90069 has finished for PR 21210 at commit 51d4c0e.

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

gatorsmile · 2018-05-02T20:55:33Z

With this fix, we still hit this test failure org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite.(It is not a test it is a sbt.testing.SuiteSelector)?

dongjoon-hyun · 2018-05-02T20:58:50Z

Please look at the log, @gatorsmile . That's a different issue instead of No such file or directory. With this patch, spark-submit is executed correctly and returned code 1. This one only reduce the degree of flakiness.

[info] org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite *** ABORTED *** (16 seconds, 491 milliseconds)
[info]   spark-submit returned with exit code 1.
[info]   Command line: './bin/spark-submit' '--name' 'prepare testing tables' '--master' 'local[2]' '--conf' 'spark.ui.enabled=false' '--conf' 'spark.master.rest.enabled=false' '--conf' 'spark.sql.warehouse.dir=/home/jenkins/workspace/SparkPullRequestBuilder@2/target/tmp/warehouse-ad028d51-4453-490f-92be-e65a3d7153d7' '--conf' 'spark.sql.test.version.index=0' '--driver-java-options' '-Dderby.system.home=/home/jenkins/workspace/SparkPullRequestBuilder@2/target/tmp/warehouse-ad028d51-4453-490f-92be-e65a3d7153d7' '/home/jenkins/workspace/SparkPullRequestBuilder@2/target/tmp/test6311626249493665939.py'

At least, Case 2. Test Result (no failures) is completely resolved by this PR.

dongjoon-hyun · 2018-05-02T21:02:33Z

According to the log, the latest one is DataNecleus.Metadata jar error.

2018-05-02 11:05:34.118 - stdout> 11:05:34.118 ERROR DataNucleus.MetaData: Error opening the Meta-Data file "jar:file:/tmp/test-spark/spark-2.0.2/jars/hive-metastore-1.2.1.spark2.jar!/package.jdo"

cloud-fan · 2018-05-03T02:44:34Z

retest this please

cloud-fan · 2018-05-03T02:44:38Z

LGTM

dongjoon-hyun · 2018-05-03T04:22:43Z

Thank you for review, @cloud-fan !

SparkQA · 2018-05-03T05:23:49Z

Test build #90090 has finished for PR 21210 at commit 51d4c0e.

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

dongjoon-hyun · 2018-05-03T06:48:38Z

Thank you for review, @HyukjinKwon . :)

cloud-fan · 2018-05-03T07:17:19Z

thanks, merging to master/2.3!

…fy the downloaded file ## What changes were proposed in this pull request? Although [SPARK-22654](https://issues.apache.org/jira/browse/SPARK-22654) made `HiveExternalCatalogVersionsSuite` download from Apache mirrors three times, it has been flaky because it didn't verify the downloaded file. Some Apache mirrors terminate the downloading abnormally, the *corrupted* file shows the following errors. ``` gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now 22:46:32.700 WARN org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite: ===== POSSIBLE THREAD LEAK IN SUITE o.a.s.sql.hive.HiveExternalCatalogVersionsSuite, thread names: Keep-Alive-Timer ===== *** RUN ABORTED *** java.io.IOException: Cannot run program "./bin/spark-submit" (in directory "/tmp/test-spark/spark-2.2.0"): error=2, No such file or directory ``` This has been reported weirdly in two ways. For example, the above case is reported as Case 2 `no failures`. - Case 1. [Test Result (1 failure / +1)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-sbt-hadoop-2.7/4389/) - Case 2. [Test Result (no failures)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-maven-hadoop-2.6/4811/) This PR aims to make `HiveExternalCatalogVersionsSuite` more robust by verifying the downloaded `tgz` file by extracting and checking the existence of `bin/spark-submit`. If it turns out that the file is empty or corrupted, `HiveExternalCatalogVersionsSuite` will do retry logic like the download failure. ## How was this patch tested? Pass the Jenkins. Author: Dongjoon Hyun <dongjoon@apache.org> Closes #21210 from dongjoon-hyun/SPARK-23489. (cherry picked from commit c9bfd1c) Signed-off-by: Wenchen Fan <wenchen@databricks.com>

dongjoon-hyun · 2018-05-03T07:52:06Z

Thank you for merging, @cloud-fan.
And thank you, all.

… should verify the downloaded file ## What changes were proposed in this pull request? This is a backport of #21210 because `branch-2.2` also faces the same failures. Although [SPARK-22654](https://issues.apache.org/jira/browse/SPARK-22654) made `HiveExternalCatalogVersionsSuite` download from Apache mirrors three times, it has been flaky because it didn't verify the downloaded file. Some Apache mirrors terminate the downloading abnormally, the *corrupted* file shows the following errors. ``` gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now 22:46:32.700 WARN org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite: ===== POSSIBLE THREAD LEAK IN SUITE o.a.s.sql.hive.HiveExternalCatalogVersionsSuite, thread names: Keep-Alive-Timer ===== *** RUN ABORTED *** java.io.IOException: Cannot run program "./bin/spark-submit" (in directory "/tmp/test-spark/spark-2.2.0"): error=2, No such file or directory ``` This has been reported weirdly in two ways. For example, the above case is reported as Case 2 `no failures`. - Case 1. [Test Result (1 failure / +1)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-sbt-hadoop-2.7/4389/) - Case 2. [Test Result (no failures)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-maven-hadoop-2.6/4811/) This PR aims to make `HiveExternalCatalogVersionsSuite` more robust by verifying the downloaded `tgz` file by extracting and checking the existence of `bin/spark-submit`. If it turns out that the file is empty or corrupted, `HiveExternalCatalogVersionsSuite` will do retry logic like the download failure. ## How was this patch tested? Pass the Jenkins. Author: Dongjoon Hyun <dongjoon@apache.org> Closes #21232 from dongjoon-hyun/SPARK-23489-2.

* [SPARK-23816][CORE] Killed tasks should ignore FetchFailures. SPARK-19276 ensured that FetchFailures do not get swallowed by other layers of exception handling, but it also meant that a killed task could look like a fetch failure. This is particularly a problem with speculative execution, where we expect to kill tasks as they are reading shuffle data. The fix is to ensure that we always check for killed tasks first. Added a new unit test which fails before the fix, ran it 1k times to check for flakiness. Full suite of tests on jenkins. Author: Imran Rashid <irashid@cloudera.com> Closes apache#20987 from squito/SPARK-23816. (cherry picked from commit 10f45bb) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> * [SPARK-24007][SQL] EqualNullSafe for FloatType and DoubleType might generate a wrong result by codegen. `EqualNullSafe` for `FloatType` and `DoubleType` might generate a wrong result by codegen. ```scala scala> val df = Seq((Some(-1.0d), None), (None, Some(-1.0d))).toDF() df: org.apache.spark.sql.DataFrame = [_1: double, _2: double] scala> df.show() +----+----+ | _1| _2| +----+----+ |-1.0|null| |null|-1.0| +----+----+ scala> df.filter("_1 <=> _2").show() +----+----+ | _1| _2| +----+----+ |-1.0|null| |null|-1.0| +----+----+ ``` The result should be empty but the result remains two rows. Added a test. Author: Takuya UESHIN <ueshin@databricks.com> Closes apache#21094 from ueshin/issues/SPARK-24007/equalnullsafe. (cherry picked from commit f09a9e9) Signed-off-by: gatorsmile <gatorsmile@gmail.com> * [SPARK-23963][SQL] Properly handle large number of columns in query on text-based Hive table ## What changes were proposed in this pull request? TableReader would get disproportionately slower as the number of columns in the query increased. I fixed the way TableReader was looking up metadata for each column in the row. Previously, it had been looking up this data in linked lists, accessing each linked list by an index (column number). Now it looks up this data in arrays, where indexing by column number works better. ## How was this patch tested? Manual testing All sbt unit tests python sql tests Author: Bruce Robbins <bersprockets@gmail.com> Closes apache#21043 from bersprockets/tabreadfix. * [MINOR][DOCS] Fix comments of SQLExecution#withExecutionId ## What changes were proposed in this pull request? Fix comment. Change `BroadcastHashJoin.broadcastFuture` to `BroadcastExchangeExec.relationFuture`: https://github.com/apache/spark/blob/d28d5732ae205771f1f443b15b10e64dcffb5ff0/sql/core/src/main/scala/org/apache/spark/sql/execution/exchange/BroadcastExchangeExec.scala#L66 ## How was this patch tested? N/A Author: seancxmao <seancxmao@gmail.com> Closes apache#21113 from seancxmao/SPARK-13136. (cherry picked from commit c303b1b) Signed-off-by: hyukjinkwon <gurwls223@apache.org> * [SPARK-23941][MESOS] Mesos task failed on specific spark app name ## What changes were proposed in this pull request? Shell escaped the name passed to spark-submit and change how conf attributes are shell escaped. ## How was this patch tested? This test has been tested manually with Hive-on-spark with mesos or with the use case described in the issue with the sparkPi application with a custom name which contains illegal shell characters. With this PR, hive-on-spark on mesos works like a charm with hive 3.0.0-SNAPSHOT. I state that this contribution is my original work and that I license the work to the project under the project’s open source license Author: Bounkong Khamphousone <bounkong.khamphousone@ebiznext.com> Closes apache#21014 from tiboun/fix/SPARK-23941. (cherry picked from commit 6782359) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> * [SPARK-23433][CORE] Late zombie task completions update all tasksets Fetch failure lead to multiple tasksets which are active for a given stage. While there is only one "active" version of the taskset, the earlier attempts can still have running tasks, which can complete successfully. So a task completion needs to update every taskset so that it knows the partition is completed. That way the final active taskset does not try to submit another task for the same partition, and so that it knows when it is completed and when it should be marked as a "zombie". Added a regression test. Author: Imran Rashid <irashid@cloudera.com> Closes apache#21131 from squito/SPARK-23433. (cherry picked from commit 94641fe) Signed-off-by: Imran Rashid <irashid@cloudera.com> * [SPARK-23489][SQL][TEST][BRANCH-2.2] HiveExternalCatalogVersionsSuite should verify the downloaded file ## What changes were proposed in this pull request? This is a backport of apache#21210 because `branch-2.2` also faces the same failures. Although [SPARK-22654](https://issues.apache.org/jira/browse/SPARK-22654) made `HiveExternalCatalogVersionsSuite` download from Apache mirrors three times, it has been flaky because it didn't verify the downloaded file. Some Apache mirrors terminate the downloading abnormally, the *corrupted* file shows the following errors. ``` gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now 22:46:32.700 WARN org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite: ===== POSSIBLE THREAD LEAK IN SUITE o.a.s.sql.hive.HiveExternalCatalogVersionsSuite, thread names: Keep-Alive-Timer ===== *** RUN ABORTED *** java.io.IOException: Cannot run program "./bin/spark-submit" (in directory "/tmp/test-spark/spark-2.2.0"): error=2, No such file or directory ``` This has been reported weirdly in two ways. For example, the above case is reported as Case 2 `no failures`. - Case 1. [Test Result (1 failure / +1)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-sbt-hadoop-2.7/4389/) - Case 2. [Test Result (no failures)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-maven-hadoop-2.6/4811/) This PR aims to make `HiveExternalCatalogVersionsSuite` more robust by verifying the downloaded `tgz` file by extracting and checking the existence of `bin/spark-submit`. If it turns out that the file is empty or corrupted, `HiveExternalCatalogVersionsSuite` will do retry logic like the download failure. ## How was this patch tested? Pass the Jenkins. Author: Dongjoon Hyun <dongjoon@apache.org> Closes apache#21232 from dongjoon-hyun/SPARK-23489-2. * [SPARK-23697][CORE] LegacyAccumulatorWrapper should define isZero correctly ## What changes were proposed in this pull request? It's possible that Accumulators of Spark 1.x may no longer work with Spark 2.x. This is because `LegacyAccumulatorWrapper.isZero` may return wrong answer if `AccumulableParam` doesn't define equals/hashCode. This PR fixes this by using reference equality check in `LegacyAccumulatorWrapper.isZero`. ## How was this patch tested? a new test Author: Wenchen Fan <wenchen@databricks.com> Closes apache#21229 from cloud-fan/accumulator. (cherry picked from commit 4d5de4d) Signed-off-by: Wenchen Fan <wenchen@databricks.com> * [SPARK-21278][PYSPARK] Upgrade to Py4J 0.10.6 This PR aims to bump Py4J in order to fix the following float/double bug. Py4J 0.10.5 fixes this (py4j/py4j#272) and the latest Py4J is 0.10.6. **BEFORE** ``` >>> df = spark.range(1) >>> df.select(df['id'] + 17.133574204226083).show() +--------------------+ |(id + 17.1335742042)| +--------------------+ | 17.1335742042| +--------------------+ ``` **AFTER** ``` >>> df = spark.range(1) >>> df.select(df['id'] + 17.133574204226083).show() +-------------------------+ |(id + 17.133574204226083)| +-------------------------+ | 17.133574204226083| +-------------------------+ ``` Manual. Author: Dongjoon Hyun <dongjoon@apache.org> Closes apache#18546 from dongjoon-hyun/SPARK-21278. (cherry picked from commit c8d0aba) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> * [SPARK-16406][SQL] Improve performance of LogicalPlan.resolve `LogicalPlan.resolve(...)` uses linear searches to find an attribute matching a name. This is fine in normal cases, but gets problematic when you try to resolve a large number of columns on a plan with a large number of attributes. This PR adds an indexing structure to `resolve(...)` in order to find potential matches quicker. This PR improves the reference resolution time for the following code by 4x (11.8s -> 2.4s): ``` scala val n = 4000 val values = (1 to n).map(_.toString).mkString(", ") val columns = (1 to n).map("column" + _).mkString(", ") val query = s""" |SELECT $columns |FROM VALUES ($values) T($columns) |WHERE 1=2 AND 1 IN ($columns) |GROUP BY $columns |ORDER BY $columns |""".stripMargin spark.time(sql(query)) ``` Existing tests. Author: Herman van Hovell <hvanhovell@databricks.com> Closes apache#14083 from hvanhovell/SPARK-16406. * [PYSPARK] Update py4j to version 0.10.7. (cherry picked from commit cc613b5) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> (cherry picked from commit 323dc3a) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> * [SPARKR] Match pyspark features in SparkR communication protocol. (cherry picked from commit 628c7b5) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> (cherry picked from commit 16cd9ac) Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com> * Keep old-style messages for AnalysisException with ambiguous references

… should verify the downloaded file ## What changes were proposed in this pull request? This is a backport of apache#21210 because `branch-2.2` also faces the same failures. Although [SPARK-22654](https://issues.apache.org/jira/browse/SPARK-22654) made `HiveExternalCatalogVersionsSuite` download from Apache mirrors three times, it has been flaky because it didn't verify the downloaded file. Some Apache mirrors terminate the downloading abnormally, the *corrupted* file shows the following errors. ``` gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now 22:46:32.700 WARN org.apache.spark.sql.hive.HiveExternalCatalogVersionsSuite: ===== POSSIBLE THREAD LEAK IN SUITE o.a.s.sql.hive.HiveExternalCatalogVersionsSuite, thread names: Keep-Alive-Timer ===== *** RUN ABORTED *** java.io.IOException: Cannot run program "./bin/spark-submit" (in directory "/tmp/test-spark/spark-2.2.0"): error=2, No such file or directory ``` This has been reported weirdly in two ways. For example, the above case is reported as Case 2 `no failures`. - Case 1. [Test Result (1 failure / +1)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-sbt-hadoop-2.7/4389/) - Case 2. [Test Result (no failures)](https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test%20(Dashboard)/job/spark-master-test-maven-hadoop-2.6/4811/) This PR aims to make `HiveExternalCatalogVersionsSuite` more robust by verifying the downloaded `tgz` file by extracting and checking the existence of `bin/spark-submit`. If it turns out that the file is empty or corrupted, `HiveExternalCatalogVersionsSuite` will do retry logic like the download failure. ## How was this patch tested? Pass the Jenkins. Author: Dongjoon Hyun <dongjoon@apache.org> Closes apache#21232 from dongjoon-hyun/SPARK-23489-2.

[SPARK-23489][SQL][TEST] HiveExternalCatalogVersionsSuite should veri…

51d4c0e

…fy the downloaded file

dongjoon-hyun mentioned this pull request May 2, 2018

[SPARK-23971] [BACKPORT-2.3] Should not leak Spark sessions across test suites #21197

Closed

HyukjinKwon approved these changes May 3, 2018

View reviewed changes

asfgit closed this in c9bfd1c May 3, 2018

dongjoon-hyun deleted the SPARK-23489 branch May 3, 2018 07:52

dongjoon-hyun mentioned this pull request May 3, 2018

[SPARK-23489][SQL][TEST][BRANCH-2.2] HiveExternalCatalogVersionsSuite should verify the downloaded file #21232

Closed

[SPARK-23489][SQL][TEST] HiveExternalCatalogVersionsSuite should verify the downloaded file #21210

[SPARK-23489][SQL][TEST] HiveExternalCatalogVersionsSuite should verify the downloaded file #21210

Uh oh!

Conversation

dongjoon-hyun commented May 2, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What changes were proposed in this pull request?

How was this patch tested?

Uh oh!

gatorsmile commented May 2, 2018

Uh oh!

SparkQA commented May 2, 2018

Uh oh!

hvanhovell commented May 2, 2018

Uh oh!

SparkQA commented May 2, 2018

Uh oh!

dongjoon-hyun commented May 2, 2018

Uh oh!

dongjoon-hyun commented May 2, 2018

Uh oh!

dongjoon-hyun commented May 2, 2018

Uh oh!

SparkQA commented May 2, 2018

Uh oh!

gatorsmile commented May 2, 2018

Uh oh!

dongjoon-hyun commented May 2, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dongjoon-hyun commented May 2, 2018

Uh oh!

cloud-fan commented May 3, 2018

Uh oh!

cloud-fan commented May 3, 2018

Uh oh!

dongjoon-hyun commented May 3, 2018

Uh oh!

SparkQA commented May 3, 2018

Uh oh!

dongjoon-hyun commented May 3, 2018

Uh oh!

cloud-fan commented May 3, 2018

Uh oh!

dongjoon-hyun commented May 3, 2018

Uh oh!

Uh oh!

dongjoon-hyun commented May 2, 2018 •

edited

Loading

dongjoon-hyun commented May 2, 2018 •

edited

Loading