What is Serverless Architecture?

Serverless architecture, the new buzzword in town, has been gaining a lot of attention from both pros and the rookies in the tech industry. You hear about serverless architecture everywhere, from conferences to meetups to blog posts — partly due to the manner in which cloud vendors have leaned into promoting it. But serverless architecture is not just about the hype; it promises the possibility of ideal business implementations, which sounds quite pleasant to IT teams and is probably light on the budget as well. In this article, we'll talk about why serverless architecture is on the rise and why you should be paying attention to the trend. 

What is serverless architecture? 

The term serverless may be somewhat misleading. Deploying a serverless architecture doesn't mean that it runs the code without servers. The name is given only because the person or a company owning the system does not have to purchase or rent servers or virtual machines to be able to run backend code. The backend computation logic will run on the vendor-maintained servers and you will be charged only for your computational usage. Here, the containers are ephemeral, which means the containers will only spin up when you run the code, scale up if needed and will shut down when the execution is finished. In other words, serverless architecture is a way of building the application without any infrastructure management.  

Why did serverless architecture arise? 

 Necessity is the mother of invention. That is, in fact, the answer to why serverless architecture is becoming important. 

Today, IT teams don’t want to spend their resources on managing and operating servers or runtimes. 

To elaborate on that idea, let’s consider a big data use case (an area where serverless is booming). In the initial days of big data technology, companies had no other option other than to buy their server hardware then install and configure a big data cluster on those servers. If you’ve ever gone through that process, you’ll understand that it is a very tedious and time-consuming process. If there was ever a need to scale up the cluster, again they have to go through the process of buying the server, installing it, and configuring the software.  

But then came the boon of cloud, where you could use infrastructure-as-a-service and at the same time eliminate the process of maintaining the server hardware. For example, you can spin up an EC2 instance in an AWS cloud and have the big data cluster installed there. But still, the installation, configuration, and maintenance of the big data cluster must be done by a developer. This is the traditional way of doing things, but moving a step in the future leads us to serverless big data architecture. This is where vendors like Databricks provide readily available and scalable big data clusters where you just have to submit your Spark code; the backend processing is completely taken care by the vendor. Here autoscaling can just be enabled just by checking a checkbox while creating a cluster. Contrast this with the scaling up of clusters in an on-premise setup — that's a nightmare to mant enterprise IT teams.

To sum it up, serverless architecture came into the picture mainly for the following three reasons:

1. Avoiding on-premise installation, configuration, and maintenance of servers. 
2. Keeping costs down, as serverless architecture follows a pay-as-you-use model. 
3. Built-in high availability and fault tolerance. 

Types of Serverless Architecture

Serverless architectures can be broadly categorized into two categories: 

1. Function As A Service (FAAS): Applications where server-side logic is still written by the application developer, but, unlike traditional architectures, it’s run in stateless compute containers that are event-triggered, ephemeral (may only last for one invocation), and fully managed by a third party. AWS Lambda is one of the popular implementations of this. 
2. Backend As A Service (BAAS):Backend-as-a-Service (BaaS) is a cloud service model in which developers outsource all the backend code or behind the scenes processes of an application so that they only have to write and maintain the frontend. BaaS vendors provide pre-written software for activities that take place on servers, such as user authentication, database management, remote updating, and push notifications (for mobile apps), as well as cloud storage and hosting. 

Serverless architecture and Talend 

Talend supports many serverless features. For example, Talend has many specific components that work with Databricks like DBFSPut and DBFSGet which helps to put and get files from Databricks file system. Also, Talend Jobs can be submitted to a Spark cluster of Databricks. 

One of the major features included with Talend with respect to serverless is building Talend Jobs as Docker images. A Docker container image is a lightweight, standalone, executable package of software that includes everything needed to run an application: code, runtime, system tools, system libraries and settings. These Docker images become containers at runtime and run the code. From Talend 7.1.1 version onwards, there is an inbuilt capability to build Docker images out of a Job.

Talend’s strategy is to help organizations progress on a journey to serverless, beginning with containers-as-a-service, to function-as-a-service, to data platform-as-a-service, for both batch and streaming. It’s designed to support all the key users within an organization, including data engineers, data scientists, data stewards, and business analysts.

Talend Cloud supports a serverless environment, enabling organizations to easily access all cloud platforms, leverage native performance; deploy built-in security, quality, and data governance; and put data into the hands of business users when they need it. Give Talend Cloud a try today.