[GitHub] carbondata pull request #2533: [CARBONDATA-2765]handle flat folder support f...

[qiuchenjian-2]

GitHub user akashrn5 opened a pull request:

    https://github.com/apache/carbondata/pull/2533

    [CARBONDATA-2765]handle flat folder support for implicit column

    Be sure to do all of the following checklist to help us incorporate
    your contribution quickly and easily:
   
     - [ ] Any interfaces changed?
     
     - [ ] Any backward compatibility impacted?
     
     - [ ] Document update required?
   
     - [ ] 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.
           
     - [ ] For large changes, please consider breaking it into sub-tasks under an umbrella JIRA.
   


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

    $ git pull https://github.com/akashrn5/incubator-carbondata impli

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

    https://github.com/apache/carbondata/pull/2533.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 #2533
   
----
commit b5fc1f061b238be3d77445e4eb6b15d273e26069
Author: akashrn5 <akashnilugal@...>
Date:   2018-07-20T07:16:32Z

    handle flat folder support for implicit column

----


---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder1/7364/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Failed with Spark 2.2.1, Please check CI http://88.99.58.216:8080/job/ApacheCarbonPRBuilder/6125/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder1/7367/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Failed with Spark 2.2.1, Please check CI http://88.99.58.216:8080/job/ApacheCarbonPRBuilder/6129/



---

[qiuchenjian-2]

Github user akashrn5 commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    retest this please


---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder1/7371/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Success with Spark 2.2.1, Please check CI http://88.99.58.216:8080/job/ApacheCarbonPRBuilder/6132/



---

[qiuchenjian-2]

Github user ravipesala commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    SDV Build Success , Please check CI http://144.76.159.231:8080/job/ApacheSDVTests/5942/



---

[qiuchenjian-2]

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

    https://github.com/apache/carbondata/pull/2533#discussion_r204214380
 
    --- Diff: core/src/main/java/org/apache/carbondata/core/indexstore/blockletindex/BlockDataMap.java ---
    @@ -642,7 +643,17 @@ private boolean addBlockBasedOnMinMaxValue(FilterExecuter filterExecuter, byte[]
           byte[][] minValue, String filePath, int blockletId) {
         BitSet bitSet = null;
         if (filterExecuter instanceof ImplicitColumnFilterExecutor) {
    -      String uniqueBlockPath = filePath.substring(filePath.lastIndexOf("/Part") + 1);
    +      String uniqueBlockPath;
    +      String blockName = filePath.substring(filePath.lastIndexOf("/") + 1);
    +      if (filePath.contains("/Fact/Part0/Segment_")) {
    --- End diff --
   
    Use PR https://github.com/apache/carbondata/pull/2503  and check `CarbonUtil.isStandardCarbonTable` instead of doing contains check


---

[qiuchenjian-2]

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

    https://github.com/apache/carbondata/pull/2533#discussion_r204214523
 
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/hive/CarbonRelation.scala ---
    @@ -173,15 +175,38 @@ case class CarbonRelation(
                 .getValidAndInvalidSegments.getValidSegments.asScala
               var size = 0L
               // for each segment calculate the size
    -          segments.foreach {validSeg =>
    -            // for older store
    -            if (null != validSeg.getLoadMetadataDetails.getDataSize &&
    -                null != validSeg.getLoadMetadataDetails.getIndexSize) {
    -              size = size + validSeg.getLoadMetadataDetails.getDataSize.toLong +
    -                     validSeg.getLoadMetadataDetails.getIndexSize.toLong
    -            } else {
    -              size = size + FileFactory.getDirectorySize(
    -                CarbonTablePath.getSegmentPath(tablePath, validSeg.getSegmentNo))
    +          if (carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    +                .get(CarbonCommonConstants.FLAT_FOLDER).isDefined &&
    +              carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    --- End diff --
   
    Why is this check required? In flat folder case also we can still get the datasize from tablestatus. I don't think these changes are required at all.


---

[qiuchenjian-2]

Github user akashrn5 commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/2533#discussion_r204282454
 
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/hive/CarbonRelation.scala ---
    @@ -173,15 +175,38 @@ case class CarbonRelation(
                 .getValidAndInvalidSegments.getValidSegments.asScala
               var size = 0L
               // for each segment calculate the size
    -          segments.foreach {validSeg =>
    -            // for older store
    -            if (null != validSeg.getLoadMetadataDetails.getDataSize &&
    -                null != validSeg.getLoadMetadataDetails.getIndexSize) {
    -              size = size + validSeg.getLoadMetadataDetails.getDataSize.toLong +
    -                     validSeg.getLoadMetadataDetails.getIndexSize.toLong
    -            } else {
    -              size = size + FileFactory.getDirectorySize(
    -                CarbonTablePath.getSegmentPath(tablePath, validSeg.getSegmentNo))
    +          if (carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    +                .get(CarbonCommonConstants.FLAT_FOLDER).isDefined &&
    +              carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    --- End diff --
   
    if `validSeg.getLoadMetadataDetails.getDataSize` or `alidSeg.getLoadMetadataDetails.getIndexSize` is null, then it will try to get size using the segment path which will not present in flat folder case, it will throw segment does not exists exception, i got this exception, so we can handle like this


---

[qiuchenjian-2]

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

    https://github.com/apache/carbondata/pull/2533#discussion_r204284556
 
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/hive/CarbonRelation.scala ---
    @@ -173,15 +175,38 @@ case class CarbonRelation(
                 .getValidAndInvalidSegments.getValidSegments.asScala
               var size = 0L
               // for each segment calculate the size
    -          segments.foreach {validSeg =>
    -            // for older store
    -            if (null != validSeg.getLoadMetadataDetails.getDataSize &&
    -                null != validSeg.getLoadMetadataDetails.getIndexSize) {
    -              size = size + validSeg.getLoadMetadataDetails.getDataSize.toLong +
    -                     validSeg.getLoadMetadataDetails.getIndexSize.toLong
    -            } else {
    -              size = size + FileFactory.getDirectorySize(
    -                CarbonTablePath.getSegmentPath(tablePath, validSeg.getSegmentNo))
    +          if (carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    +                .get(CarbonCommonConstants.FLAT_FOLDER).isDefined &&
    +              carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    --- End diff --
   
    No, it should be present in case of flat folder case also. If not present then fix adding the datasize to tablestatus not this way


---

[qiuchenjian-2]

Github user akashrn5 commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/2533#discussion_r204284734
 
    --- Diff: core/src/main/java/org/apache/carbondata/core/indexstore/blockletindex/BlockDataMap.java ---
    @@ -642,7 +643,17 @@ private boolean addBlockBasedOnMinMaxValue(FilterExecuter filterExecuter, byte[]
           byte[][] minValue, String filePath, int blockletId) {
         BitSet bitSet = null;
         if (filterExecuter instanceof ImplicitColumnFilterExecutor) {
    -      String uniqueBlockPath = filePath.substring(filePath.lastIndexOf("/Part") + 1);
    +      String uniqueBlockPath;
    +      String blockName = filePath.substring(filePath.lastIndexOf("/") + 1);
    +      if (filePath.contains("/Fact/Part0/Segment_")) {
    --- End diff --
   
    i was thinking of adding a utility function, but here we will not get the carbon table, so i thought i will use this way


---

[qiuchenjian-2]

Github user akashrn5 commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/2533#discussion_r204291557
 
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/hive/CarbonRelation.scala ---
    @@ -173,15 +175,38 @@ case class CarbonRelation(
                 .getValidAndInvalidSegments.getValidSegments.asScala
               var size = 0L
               // for each segment calculate the size
    -          segments.foreach {validSeg =>
    -            // for older store
    -            if (null != validSeg.getLoadMetadataDetails.getDataSize &&
    -                null != validSeg.getLoadMetadataDetails.getIndexSize) {
    -              size = size + validSeg.getLoadMetadataDetails.getDataSize.toLong +
    -                     validSeg.getLoadMetadataDetails.getIndexSize.toLong
    -            } else {
    -              size = size + FileFactory.getDirectorySize(
    -                CarbonTablePath.getSegmentPath(tablePath, validSeg.getSegmentNo))
    +          if (carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    +                .get(CarbonCommonConstants.FLAT_FOLDER).isDefined &&
    +              carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    --- End diff --
   
    ok


---

[qiuchenjian-2]

Github user akashrn5 commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/2533#discussion_r204297035
 
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/hive/CarbonRelation.scala ---
    @@ -173,15 +175,38 @@ case class CarbonRelation(
                 .getValidAndInvalidSegments.getValidSegments.asScala
               var size = 0L
               // for each segment calculate the size
    -          segments.foreach {validSeg =>
    -            // for older store
    -            if (null != validSeg.getLoadMetadataDetails.getDataSize &&
    -                null != validSeg.getLoadMetadataDetails.getIndexSize) {
    -              size = size + validSeg.getLoadMetadataDetails.getDataSize.toLong +
    -                     validSeg.getLoadMetadataDetails.getIndexSize.toLong
    -            } else {
    -              size = size + FileFactory.getDirectorySize(
    -                CarbonTablePath.getSegmentPath(tablePath, validSeg.getSegmentNo))
    +          if (carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    +                .get(CarbonCommonConstants.FLAT_FOLDER).isDefined &&
    +              carbonTable.getTableInfo.getFactTable.getTableProperties.asScala
    --- End diff --
   
    i have handled it, as you said, no need of these changes


---

[qiuchenjian-2]

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

    https://github.com/apache/carbondata/pull/2533#discussion_r204316346
 
    --- Diff: core/src/main/java/org/apache/carbondata/core/indexstore/blockletindex/BlockDataMap.java ---
    @@ -642,7 +643,17 @@ private boolean addBlockBasedOnMinMaxValue(FilterExecuter filterExecuter, byte[]
           byte[][] minValue, String filePath, int blockletId) {
         BitSet bitSet = null;
         if (filterExecuter instanceof ImplicitColumnFilterExecutor) {
    -      String uniqueBlockPath = filePath.substring(filePath.lastIndexOf("/Part") + 1);
    +      String uniqueBlockPath;
    +      String blockName = filePath.substring(filePath.lastIndexOf("/") + 1);
    +      if (filePath.contains("/Fact/Part0/Segment_")) {
    --- End diff --
   
    `CarbonTable` is available in BlockletDataMapModel, so you can do the check.


---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Success with Spark 2.1.0, Please check CI http://136.243.101.176:8080/job/ApacheCarbonPRBuilder1/7390/



---

[qiuchenjian-2]

Github user CarbonDataQA commented on the issue:

    https://github.com/apache/carbondata/pull/2533
 
    Build Failed with Spark 2.2.1, Please check CI http://88.99.58.216:8080/job/ApacheCarbonPRBuilder/6151/



---

[qiuchenjian-2]

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

    https://github.com/apache/carbondata/pull/2533#discussion_r204331739
 
    --- Diff: core/src/main/java/org/apache/carbondata/core/indexstore/blockletindex/BlockDataMap.java ---
    @@ -97,10 +98,15 @@
        * partition table and non transactional table
        */
       protected boolean isFilePathStored;
    +  /**
    +   * datamap model
    +   */
    +  protected BlockletDataMapModel blockletDataMapModel;
    --- End diff --
   
    Don't take complete `blockletDataMapModel` to class level, it wastes memory as we keep them in LRU . Just keep boolean isStandardCarbonTable in class level


---