Category: Spark

The Hadoop in Real World team explains the notion of broadcast variables in Apache Spark:

Broadcast variables are variables which are available in all executors executing the Spark application. These variables are already cached and ready to be used by tasks executing as part of the application. Broadcast variables are sent to the executors only once and it is available for all tasks executing in the executors.

Read on to understand when they are useful and, just as importantly, when not to use them. They seem like the type of thing which a newer developer could easily misuse.

The Hadoop in Real World team shows how to deduplicate rows in a DataFrame in Spark:

It is a pretty common use case to find the list of duplicate elements or rows in a Spark DataFrame and it is very easy to do with a groupBy() and a count()

Where “easy” has as a modifier just how many columns you’re dealing with in the DataFrame.

The Hadoop in Real World team compares two functions against RDDs in Spark:

Let’s examine the below aggregateByKey. The first parameter – 0 is the initial value and also indicates the type of the output.

First _+_  function indicates the function on the map side combine and second _+_ function indicates the reduce side combine. Both functions are the same in this case.

This is a demo-driven post, so check it out.

David Greenshtein shows how we can build type-2 slowly changing dimensions using Apache Hudi:

Implementing SCD2 in a data lake without using an additional framework like Apache Hudi introduces the challenge of updating data stored on immutable Amazon S3 storage, and as a result requires the implementor to create multiple copies of intermediate results. This situation may lead to a significant maintenance effort and potential data loss or data inconsistency.

Apache Hudi is an open-source data management framework used to simplify incremental data processing and data pipeline development. Hudi enables you to manage data at the record level in Amazon S3 and helps to handle data privacy use cases requiring record-level updates and deletes. Hudi is supported by Amazon EMR starting from version 5.28 and is automatically installed when you choose Spark, Hive, or Presto when deploying your EMR cluster.

Click through for the process.

Landon Robinson shows how to sort data in Spark RDDs and DataFrames:

In the analysis section of Spark Starter Guide 4.6: How to Aggregate Data, we asked these questions: “Who is the youngest cat in the data? Who is the oldest?”

Let’s use ordering in Spark as an alternative method to answer those same questions, and achieve the same result. Specifically, let’s again find the youngest and oldest cats in the data.

Click through for plenty of examples.