Apache Beam | A quick guide

Reading Time: 5 minutes

What is Apache Beam?

Apache Beam is a product of the Apache Software Foundation. Which is an integrated open-source programming model used to define and use data processing pipelines. Including ETL, namely, ETL, Extract, Transform, Load, and both data collection and distribution. we have used Python and Java two editing languages for this model.

The first version of Apache Beam was released in June 2016 and its stable release (2.7.0) it was made available in October 2018. With the help of any open source Beam SDKs, we can create a system that can define the pipeline. Thereafter, this pipeline will be used by any distributed Beam for background processing that includes other beam runners such as –

Why Apache Beam is important?

The reason Beam exists is that it can be used in similar data processing tasks that can be subdivided into a few smaller pieces of data that are processed independently.it can convert a set of data of any size even if the input data is limited i.e., from a bulk data source or an endless data set from a distributed data source this is the advantage of using a beam SDK. Another advantage of using the Beam SDK is that it uses the same section to display both bound and unlimited data. Currently, Beam supports the following SDKs in a specific language –

While creating a beam pipeline, one can have the following processing tasks in terms of abstractions –

  • Pipeline
  • PCollection
  • PTransform

Pipeline The pipe contains all the data processing functions from start to finish. The sections involved in this are reading the input data, converting that data, and after that, writing the output. When we build a pipe, we have to provide an extraction option, which tells the pipe where it ran and how it worked.

PCollection As the name suggests, it represents a set of distributed data on which a line pipeline should operate. The data set can be restricted or unrestricted, it can come from a fixed source or it can come from a continuous source that updates with the help of registration or any other method.

PTransform Represents data conversion or functionality. The integration of the entire PTransform is PCollection, performs the processing functions we provide, and provides egg or more PCollection components such as output.

What is the Pipeline Development Life-cycle of Apache Beam?

There exist three significant stages while designing a pipeline, and that is as follows:

Developing a pipeline – Designing a beam pipeline involves the information about how to determine a pipeline structure. how to choose that, which transform must apply to the data, and how to determine the input and output data. Following are the things while designing a pipeline –

  • Where the input data is stored i.e., how many sets of input data it has, it also determines which kind of read transforms we will need to apply at the beginning of the pipeline.
  • What does the input data look like i.e., the data can be plain text or rows in a database table?
  • What to do with the data, involves how we change and manipulate our data and determine how to build core transforms such as ParDo.
  • Where should the data go i.e., what kind of write transforms we will need to apply at the end of the pipeline.

Apache Beam Structure for Basic Pipeline

The most manageable pipeline would represent a linear flow of operations. A pipeline can have multiple input sources, as well as numerous output sinks. While creating a beam pipeline, one can have the following processing tasks in terms of abstractions –

Creating a Pipeline

For constructing a pipeline with the help of classes in the beam SDKs. The program will need to perform the following steps –

  • Create a pipeline object – Each pipeline object is an independent entity in beam SDK. That takes both the data and the transforms which are to be applied to the data.
  • Read data into the pipeline – For reading the data, there are two root transforms, and that is Read and Create. With the help of Reading transform, we understand the data from an external source like a text file, and with the help of Create transform, we create a PCollection from in-memory.
  • Apply transforms to process pipeline data – With the help of application method on each PCollection, we can handle our data using various transforms in the beam SDKs.
  • Output the final transformed PCollections – it needs to output the data after the pipeline has been applied all of its transforms. Forgiving the output of final PCollection, we refer to write transform, and it would output the elements of a PCollection to an external database.
  • Run the pipeline – For executing our pipeline, we use the run method. If we would like to block the execution, then run the pipeline by appending a technique known as WaitUntilFinish.
  • Testing a pipeline – Testing is one of the most essential parts of building a pipeline, before running the pipeline on the runners, we should perform unit testing on it. After testing it choose a small runner like Flink and run the pipeline on it. For the support of Unit testing, the beam SDKs for Java provide many test classes in the testing package. Some of the testing objects are –
  • Testing Individual DoFn Objects – In Beam SDK for Java gives way for testing. A DoFn, which is known as DoFn Tester. Which uses the JUnit framework. This is included in the SDK Transforms package in Java.
  • Testing Composite Transforms – For testing a composite transform that we have created. First, we have created a pipeline and some known input data. After that, apply the composite transform to the input data set or PCollection and save the output PCollection.
  • Testing a pipeline end to end – For testing a pipeline end to end. A class is already provided in the Beam. Java SDK which is Test Pipeline. When we create a pipeline object in place of the pipeline we can use it.

Branching PCollection

As transforms do not consume PCollection. But take each element of the PCollection and with the help of each component. And it creates a new PCollection as an output. The advantage of consuming each aspect is that we can perform different tasks to different elements in the same PCollection.

Merging PCollections

For merging two or more PCollection in a single unit. There are two transforms or methods available, and that is as follows :

Flatten – For merging to or more PCollections of the same type we can use beam SDK. 

Join – For performing a relational join between two or more PCollections. We can use CoGroupByKey transform. The only condition for using this transform is that the PCollections must be in key, value pair, and of the same type.

A Strategic Approach for Distributed Processing

Apache Beam can be implemented in several use cases. One of the examples of Beam is a mobile game, which has some complex functionalities. And with the help of several pipelines, the complex tasks can be easily implemented. In this example, there exists a mobile game, and several users can play it at a time from different locations. Users can also play in teams sitting at various locations. So the data generated will be in a considerable amount due to which the event time and the processing time of the scores of different teams can vary. And this time difference between the processing time and event time is called Skew. So with the help of pipelines available in Apache Beam Architecture. We can reduce this time difference, and the scores will not vary.

References :

Written by 

Udit is a Software Consultant at Knoldus . He has completed his Masters of Computer Applications from Vellore institute of Technology. He is enthusiastic ,hard-working and determine person with strong attention to detail and eager to learn about new technologies.