AWS Lambda — most widely adopted uses cases
Serverless computing unmoving remains a Buzz in the digital disruption space, where almost all enterprises desired to have their Application architecture reengineered to Micro services and leverage the advantages of Serverless computing.
Whenever Serverless is being discussed, thanks to AWS for services like Lambda, this future-forward trend in application development and cloud-based hosting is witnessing great advancements.
For those who still aren’t aware about AWS Lambda, I strongly recommend them to go through the AWS documentation for Lambda Service. In simple explanation, AWS Lambda is Amazon Serverless compute service that allows you to run code without having to manage the servers. So in nutshell, it enables developers to focus only on code in building the world class applications and worry less about the infrastructure management.
In the following article, I will cover up the most widely adopted enterprise uses cases for Lambda in the industry.
1) Data Lakes
A cloud-native data warehouse as a replacement of a traditional legacy data warehouse. Serverless data lakes on AWS offers a great flexibility and agility that was missing from traditional data warehouse.
With the flexibility we have with the Serverless data lakes these days, enterprises are geared up for any kind of analytics such as fraud detection, recommendations, insights, event-driven ETL, IOTs etc. Also, it enables enterprises with the most powerful flexibility of scalability.
A data lake is the central repository for all data in the enterprise. It ingests data from a variety of data sources and stores them in the lake in the format they arrive without losing data. ETL processes will transform the data into a prepared state, either as files optimized for query such as ORC or parquet, or into Redshift, an MPP Data warehouse with many nodes.
Traditional data warehouses used to store data on rational databases that store both the data and perform operations as well. This leads to a challenge of large data grow and performance issues.
Today enterprises are leveraging the cloud-native solutions for data lakes such as Amazon which enables them a great flexibility and scalability as a combination. With Amazon S3, enterprises have a cost-effectiveness that can store virtually unlimited amount of data and enables organization to define automation rules around their data.
2) Microservices
An Architectural and organizational approach to a software development where software is further composed into independent services that communicate over well-defined APIs. Traditionally, all such services were not loosely coupled but where tightly in nature and run as a single service.
With Microservices architecture, application is built as independent components that run each application process as a service. The services communicate over a well-defined interface using APIs. Every single service is responsible to perform a single function and run independently.
“No server is easier to manage than no server”
In above figure 2 shows the architecture pattern of a microservice with AWS Lambda where the complete service is built on managed service and eliminates the architecture burden to design for scale and high availability and eliminates the operational efforts of running and monitoring the microservice underlying infrastructure.
In the example above, we have a user interface which is hosted on Amazon S3 with Cloud Front integration for caching layer. Whereas Microservices are hosted on API Gateway as individual service which enables and manages all concurrency and computing needs on its own. Last layer talks about Data layer where using the Serverless service Lambda is performing CRUD function on data layer.
3) Disaster Recovery
Many enterprises are using Lambda an event driven architecture to automate monitoring needs and performing the auto heal actions. This has brought out a tremendous improvement in disaster recovery from enterprises with an easy.
Consider AWS CloudWatch alarms as an AWS Lambda assisted disaster recovery use case. Where these alarms scan all EC2 instances and filter them with a particular predefined tag and even if you have hundreds of servers, a single Lambda function can utilize a thousand automated monitors and scan all EC2 instances. Lambda will also automatically remove all those monitors when no longer needed, saving you unnecessary costs.
Let’s take an example walkthrough where stopping non-production servers overnight is needed to have a cost optimized architecture which does not incur cost to enterprise.
Organizations create a Lambda function which performs a stop/start event to specific tagged EC2 instances. Now, remember Lambda needs an event to trigger a function and in this case, Cloud watch will be enabled to trigger a rule that will call Lambda function at night time.
Conclusion
In the following article I have just summarized few simple most common use cases with Serverless using AWS Lambda. There can many more such use cases which are different from enterprise to enterprise.
Serverless architecture is way more convenient and manageable.
