[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364531550 ## File path: docs/query-with-spark-sql-performance -tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the number of columns applied SUM operator reaches a certain threshold, the query time increases dramatically. we define "counter" for the meaning of columns applied SUM operator. Below will use "counter" instead of the columns applied SUM operator. + +As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of counter, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of counter exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to ["spark.sql.codegen.wholeStage"](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html), which is enabled by default since spark 2.0. and it will do all the *internal optimization possible from the spark catalyst side*. + +**Whole-Stage Java Code Generation** (aka *Whole-Stage CodeGen*) is a physical query optimization in Spark SQL that fuses multiple physical operators (as a subtree of plans that [support code generation](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-CodegenSupport.html)) together into a single Java function. + +Whole-Stage Java Code Generation improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. + +When counter is too large, the logic calculation function of hashaggregate operator is too large (there are nearly 3000 lines of code when there are 34 counters). Java itself does not recommend too large methods, which will reduce the processing efficiency, and exceed the JIT threshold, making the final execution in the way of interpretation. When the query grows in the counter, the performance drops sharply. Review comment: okay This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364233771 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. Review comment: done This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364223921 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of sum exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to spark.sql.codegen.wholeStage, which is enabled by default for spark 2.0. and it will do all the *internal optimization possible from the spark catalist side*. [https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html) + +**Whole-Stage Java Code Generation** (aka *Whole-Stage CodeGen*) is a physical query optimization in Spark SQL that fuses multiple physical operators (as a subtree of plans that [support code generation](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-CodegenSupport.html)) together into a single Java function. + +Whole-Stage Java Code Generation improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. + +When sum is too large, the logic calculation function of hashaggregate operator is too large (there are nearly 3000 lines of code when there are 34 indicators). Java itself does not recommend too large methods, which will reduce the processing efficiency, and exceed the JIT threshold, making the final execution in the way of interpretation. When the carbon grows in the counter, the performance drops sharply. Review comment: yes, modify. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364223612 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of sum exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to spark.sql.codegen.wholeStage, which is enabled by default for spark 2.0. and it will do all the *internal optimization possible from the spark catalist side*. [https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html) Review comment: done This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364223369 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning Review comment: sorry,both modify file name and title. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364219468 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of sum exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to spark.sql.codegen.wholeStage, which is enabled by default for spark 2.0. and it will do all the *internal optimization possible from the spark catalist side*. [https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html) + +**Whole-Stage Java Code Generation** (aka *Whole-Stage CodeGen*) is a physical query optimization in Spark SQL that fuses multiple physical operators (as a subtree of plans that [support code generation](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-CodegenSupport.html)) together into a single Java function. + +Whole-Stage Java Code Generation improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. + +When sum is too large, the logic calculation function of hashaggregate operator is too large (there are nearly 3000 lines of code when there are 34 indicators). Java itself does not recommend too large methods, which will reduce the processing efficiency, and exceed the JIT threshold, making the final execution in the way of interpretation. When the carbon grows in the counter, the performance drops sharply. + +But unfortunately, spark 2.1 only provides switching capability. User can only choose to turn the function on or off. This leads to a sharp drop in performance when the aggregation operator is too large. + +Fortunately, spark 2.3 provide more configuration. Users can better configure this parameter. + +So in spark 2.1, when the number of operators cannot be confirmed, spark.sql.codegen.wholeStage can be turned off to ensure the query efficiency. when in spark 2.3 and above users can open spark.sql.codegen.wholeStage and configure it. Review comment: It's reasonable. I'll adjust it. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364216102 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of sum exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to spark.sql.codegen.wholeStage, which is enabled by default for spark 2.0. and it will do all the *internal optimization possible from the spark catalist side*. [https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html) + +**Whole-Stage Java Code Generation** (aka *Whole-Stage CodeGen*) is a physical query optimization in Spark SQL that fuses multiple physical operators (as a subtree of plans that [support code generation](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-CodegenSupport.html)) together into a single Java function. + +Whole-Stage Java Code Generation improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. + +When sum is too large, the logic calculation function of hashaggregate operator is too large (there are nearly 3000 lines of code when there are 34 indicators). Java itself does not recommend too large methods, which will reduce the processing efficiency, and exceed the JIT threshold, making the final execution in the way of interpretation. When the carbon grows in the counter, the performance drops sharply. + +But unfortunately, spark 2.1 only provides switching capability. User can only choose to turn the function on or off. This leads to a sharp drop in performance when the aggregation operator is too large. + +Fortunately, spark 2.3 provide more configuration. Users can better configure this parameter. Review comment: add link for user to check the configurations. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364211975 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of sum exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to spark.sql.codegen.wholeStage, which is enabled by default for spark 2.0. and it will do all the *internal optimization possible from the spark catalist side*. [https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html) + +**Whole-Stage Java Code Generation** (aka *Whole-Stage CodeGen*) is a physical query optimization in Spark SQL that fuses multiple physical operators (as a subtree of plans that [support code generation](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-CodegenSupport.html)) together into a single Java function. + +Whole-Stage Java Code Generation improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. + +When sum is too large, the logic calculation function of hashaggregate operator is too large (there are nearly 3000 lines of code when there are 34 indicators). Java itself does not recommend too large methods, which will reduce the processing efficiency, and exceed the JIT threshold, making the final execution in the way of interpretation. When the carbon grows in the counter, the performance drops sharply. Review comment: yes, your suggestion is good, has modify. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364208450 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): + +![File Directory Structure](../docs/images/codegen.png?raw=true) + +The horizontal axis is the number of sum, and the vertical axis is the time consumed in seconds. + +It can be seen from the figure that when the number of sum exceeds 85, the query time is significantly increased. + +After analysis, this problem is related to spark.sql.codegen.wholeStage, which is enabled by default for spark 2.0. and it will do all the *internal optimization possible from the spark catalist side*. [https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/spark-sql-whole-stage-codegen.html) Review comment: okay, done This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364196990 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) + +## The influence of spark.sql.codegen.wholeStage configuration on query + +In practice, we found that when the sum of CarbonData's queries reaches a certain threshold, the query time increases dramatically. As shown in the figure below(spark 2.1): Review comment: okay,done This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r364193833 ## File path: docs/query-with-spark-sql-performacne-tuning.md ## @@ -0,0 +1,58 @@ + + +# Query with spark-sql performacne tuning + This tutorial guides you to create CarbonData Tables and optimize performance. + The following sections will elaborate on the below topics : + + * [The influence of spark.sql.codegen.wholeStage configuration on query](#The influence of spark.sql.codegen.wholeStage configuration on query) Review comment: manhua,need press ctrl then can jump to the details. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r362229446 ## File path: docs/performance-tuning.md ## @@ -173,6 +173,8 @@ | carbon.sort.temp.compressor | spark/carbonlib/carbon.properties | Data loading | Specify the name of compressor to compress the intermediate sort temporary files during sort procedure in data loading. | The optional values are 'SNAPPY','GZIP','BZIP2','LZ4','ZSTD', and empty. Specially, empty means that Carbondata will not compress the sort temp files. This parameter will be useful if you encounter disk bottleneck. | | carbon.load.skewedDataOptimization.enabled | spark/carbonlib/carbon.properties | Data loading | Whether to enable size based block allocation strategy for data loading. | When loading, carbondata will use file size based block allocation strategy for task distribution. It will make sure that all the executors process the same size of data -- It's useful if the size of your input data files varies widely, say 1MB to 1GB. | | carbon.load.min.size.enabled | spark/carbonlib/carbon.properties | Data loading | Whether to enable node minumun input data size allocation strategy for data loading.| When loading, carbondata will use node minumun input data size allocation strategy for task distribution. It will make sure the nodes load the minimum amount of data -- It's useful if the size of your input data files very small, say 1MB to 256MB,Avoid generating a large number of small files. | +| spark.sql.codegen.wholeStage | spark/conf/spark-defaults.conf | Querying | improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. | The whole stage CodeGen mechanism introduced by spark SQL in version 2. X causes. This configuration is recommended to be off at spark 2.1 and on at spark 2.3. Because under spark2.1 user can only use spark.sql.codegen.wholeStage to control whether to use codegen, but can not config the size of the method. In fact, this parameter should be configured to be the same as the local JDK. Under spark2.3 support spark.sql.codegen.hugeMethodLimit use can use that to config the method size. | Review comment: i has move spark sql codegen optimization guid line to a new md file named query-with-spark-sql-performacne-tuning. please check. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [carbondata] MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support
MarvinLitt commented on a change in pull request #3518: [DOC] add performance-tuning with codegen parameters support URL: https://github.com/apache/carbondata/pull/3518#discussion_r361905323 ## File path: docs/performance-tuning.md ## @@ -173,6 +173,8 @@ | carbon.sort.temp.compressor | spark/carbonlib/carbon.properties | Data loading | Specify the name of compressor to compress the intermediate sort temporary files during sort procedure in data loading. | The optional values are 'SNAPPY','GZIP','BZIP2','LZ4','ZSTD', and empty. Specially, empty means that Carbondata will not compress the sort temp files. This parameter will be useful if you encounter disk bottleneck. | | carbon.load.skewedDataOptimization.enabled | spark/carbonlib/carbon.properties | Data loading | Whether to enable size based block allocation strategy for data loading. | When loading, carbondata will use file size based block allocation strategy for task distribution. It will make sure that all the executors process the same size of data -- It's useful if the size of your input data files varies widely, say 1MB to 1GB. | | carbon.load.min.size.enabled | spark/carbonlib/carbon.properties | Data loading | Whether to enable node minumun input data size allocation strategy for data loading.| When loading, carbondata will use node minumun input data size allocation strategy for task distribution. It will make sure the nodes load the minimum amount of data -- It's useful if the size of your input data files very small, say 1MB to 256MB,Avoid generating a large number of small files. | +| spark.sql.codegen.wholeStage | spark/conf/spark-defaults.conf | Querying | improves the execution performance of a query by collapsing a query tree into a single optimized function that eliminates virtual function calls and leverages CPU registers for intermediate data. | The whole stage CodeGen mechanism introduced by spark SQL in version 2. X causes. This configuration is recommended to be off at spark 2.1 and on at spark 2.3. Because under spark2.1 user can only use spark.sql.codegen.wholeStage to control whether to use codegen, but can not config the size of the method. In fact, this parameter should be configured to be the same as the local JDK. Under spark2.3 support spark.sql.codegen.hugeMethodLimit use can use that to config the method size. | Review comment: Some spark configurations are helpful for query performance improvement. Can we add a chapter in FAQ or an MD file to record these parameters? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services