I read that in Apache Spark, GroupBy is a wide transformation meaning it requires data shuffle.My doubt is : Let's say I do df.groupBy(column).count(), so will the partitions first groupBy and count the values within their own partition and then share the result with the other partitions or will it be the case where data for similar keys are transfered to a common partition and then count operation would take place on each partition?
Is there spark conf property available to default the number of partitions after UnionAll operation in Spark.. In case joins and aggregations, spark.sql.shuffle.partitions value is used as default partition size and do we have similar property to restrict the number of partitions after UnionAll operation.. The problem which I see now is if I join dataframe df1 to df2, the number of resulting partition is df1.partitions + df2.partitions and I am looking for a solution to restrict the number of resulting partitions of all unions in my program..
I am using repartition on columns to store the data in parquet. But
I see that the no. of parquet partitioned files are not same with the
no. of Rdd partitions. Is there no correlation between rdd partitions
and parquet partitions?
When I write the data to parquet partition and I use Rdd
repartition and then I read the data from parquet partition , is
there any condition when the rdd partition numbers will be same
during read / write?
How is bucketing a dataframe using a column id and repartitioning a
dataframe via the same column id different?
While considering the performance of joins in Spark should we be
looking at bucketing or repartitioning (or maybe both)
Couple of things here that you;re asking - Partitioning, Bucketing and Balancing of data,
Partitioning:
Partitioning data is often used for distributing load horizontally, this has performance benefit, and helps in organizing data in a logical fashion.
Partitioning tables changes how persisted data is structured and will now create subdirectories reflecting this partitioning structure.
This can dramatically improve query performance, but only if the partitioning scheme reflects common filtering.
In Spark, this is done by df.write.partitionedBy(column*) and groups data by partitioning columns into same sub directory.
Bucketing:
Bucketing is another technique for decomposing data sets into more manageable parts. Based on columns provided, the entire data is hashed into a user-defined number of buckets (files).
Synonymous to Hive's Distribute By
In Spark, this is done by df.write.bucketBy(n, column*) and groups data by partitioning columns into same file. number of files generated is controlled by n
Repartition:
It returns a new DataFrame balanced evenly based on given partitioning expressions into given number of internal files. The resulting DataFrame is hash partitioned.
Spark manages data on these partitions that helps parallelize distributed data processing with minimal network traffic for sending data between executors.
In Spark, this is done by df.repartition(n, column*) and groups data by partitioning columns into same internal partition file. Note that no data is persisted to storage, this is just internal balancing of data based on constraints similar to bucketBy
Tl;dr
1) I am using repartition on columns to store the data in parquet. But I see that the no. of parquet partitioned files are not same with the no. of Rdd partitions. Is there no correlation between rdd partitions and parquet partitions?
repartition has correlation to bucketBy not partitionedBy. partitioned files is governed by other configs like spark.sql.shuffle.partitions and spark.default.parallelism
2) When I write the data to parquet partition and I use Rdd repartition and then I read the data from parquet partition , is there any condition when the rdd partition numbers will be same during read / write?
during read time, the number of partitions will be equal to spark.default.parallelism
3) How is bucketing a dataframe using a column id and repartitioning a dataframe via the same column id different?
Working similar, except, bucketing is a write operation and is used for persistence.
4) While considering the performance of joins in Spark should we be looking at bucketing or repartitioning (or maybe both)
repartition of both datasets are in memory, if one or both the datasets are persisted, then look into bucketBy also.
How can I calculate the trade off between number of partitions and size of DataFrame in Spark whit spark.conf.set configuration?
Example - Now assume we have an input RDD input which is filtered in the second step. Now I want to calculate the data size in the filtered RDD and calculate how many partitions will be required to repartition by considering block size is 128MB
This will help me out to pass the number of partitions to repartition method.
InputRDD=sc.textFile("sample.txt")
FilteredRDD=InputRDD.Filter( Some Filter Condition )
FilteredRDD.repartition(XX)
Q1.How to calculate the value of XX ?
Q2.What is the similar approach for Spark SQL/DataFrame?
The block size of 128MB will comes into picture only when reading /writing the data from/to HDFS. Once RDD is created, data is in memory or spill to disk based on executor RAM size.
You can't calculate data size unless calling collect() action on filtered RDD and it is not recommended.
The maximum partition size is 2GB, you can choose the number of partition based on cluster size or data model.
df.partition(col)