[GitHub] carbondata pull request #2796: [CARBONDATA-2991]NegativeArraySizeException d...

[qiuchenjian-2]

GitHub user BJangir opened a pull request:

    https://github.com/apache/carbondata/pull/2796

    [CARBONDATA-2991]NegativeArraySizeException during query execution

    Issue :- During Query Execution sometime NegativeArraySizeException  Exception in Some Tasks . And sometime Executor is lost (JVM crash)
   
    Root Cause :- It is because existing memoryblock is removed while it was in-use. This happened  because duplicate taskid generated. Sometime freed same memory addresses are assigned to another task which will initialize memory block to0 and this cause NegativeSizeArrayException whereas sometime freed memory will not be used any task of executor process but running task will try to access it and as that address is not part of process so JVM crash will happen.
   
    Solution :-  Change taskID generation to UUID based instead of System.nanoTime()
   
    Be sure to do all of the following checklist to help us incorporate
    your contribution quickly and easily:
   
     - [ ] Any interfaces changed?
     NA
     - [ ] Any backward compatibility impacted?
     NA
     - [ ] Document update required?
    NA
     - [ ] Testing done
            Please provide details on
            - Whether new unit test cases have been added or why no new tests are required?
            - How it is tested? Please attach test report.
            - Is it a performance related change? Please attach the performance test report.
            - Any additional information to help reviewers in testing this change.
         
    Manual Testing Done  
     - [ ] For large changes, please consider breaking it into sub-tasks under an umbrella JIRA.
    NA


You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/BJangir/incubator-carbondata CARBONDATA-2991

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/carbondata/pull/2796.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #2796
   
----
commit b41b35f9fdd86e7080e2d03d3219e260bcad7d31
Author: BJangir <babulaljangir111@...>
Date:   2018-10-03T17:35:42Z

    [CARBONDATA-2991]NegativeArraySizeException during query execution

----


---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder2.1/690/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    Build Success with Spark 2.3.1, Please check CI http://136.243.101.176:8080/job/carbondataprbuilder2.3/8955/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    Build Success with Spark 2.2.1, Please check CI http://95.216.28.178:8080/job/ApacheCarbonPRBuilder1/887/



---

[qiuchenjian-2]

Github user sujith71955 commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    @BJangir @jackylk @chenliang613 @ravipesala
    I think UUID.randomUUID().toString() has some performance bottle necks even though its negligible, as it internally uses SecureRandom API which is a synchronized method.
    if possible explore more on this as our system is more performance driven.
    What about using JUG?
    There is also a time bound uuid generation which seems to work lot faster compare to normal UUID generation supported by java.
    Following link will give you a fair idea regarding implementation and performance comparison.
    http://www.dcalabresi.com/blog/java/generate-java-uuid-performance/


---

[qiuchenjian-2]

Github user ravipesala commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/2796#discussion_r222637804
 
    --- Diff: core/src/main/java/org/apache/carbondata/core/util/ThreadLocalTaskInfo.java ---
    @@ -30,7 +32,7 @@ public static void setCarbonTaskInfo(CarbonTaskInfo carbonTaskInfo) {
       public static CarbonTaskInfo getCarbonTaskInfo() {
         if (null == threadLocal.get()) {
           CarbonTaskInfo carbonTaskInfo = new CarbonTaskInfo();
    -      carbonTaskInfo.setTaskId(System.nanoTime());
    +      carbonTaskInfo.setTaskId(UUID.randomUUID().toString());
    --- End diff --
   
    Please move the generation of UUID to some UUID to some common utility so all methods can use same. In future if we need to change we can change at one place


---

[qiuchenjian-2]

Github user ravipesala commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    @sujith71955 Carbondata usage of UUID is very minimal and also it is not safe to depend on another third party dependency (non tested) without much value add.


---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder2.1/712/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder2.1/713/



---

[qiuchenjian-2]

Github user ravipesala commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    LGTM


---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2796
 
    Build Success with Spark 2.2.1, Please check CI http://95.216.28.178:8080/job/ApacheCarbonPRBuilder1/909/



---

[qiuchenjian-2]

Github user asfgit closed the pull request at:

    https://github.com/apache/carbondata/pull/2796


---