What is serverless computing?
Serverless Computing promises cost-effective cloud service delivery
Serverless computing takes the guesswork out of managing cloud infrastructure. This lets app developers focus on writing code, not managing the servers the code runs on. Serverless computing promises more cost-effective management of the underlying hardware that enables cloud apps and services to operate.
“A serverless approach offers developers, teams, and organizations a level of abstraction that enables them to minimize the time and resources invested in infrastructure management. Every component of an application benefits from this approach, from computing and the database engine to messaging, analytics, and AI. Using an end-to-end serverless platform that provides a comprehensive set of serverless technologies is the best way to ensure that the organization gains the maximum benefit from going serverless,” said Microsoft.
“The pub/sub model of communication makes it easy to ingest and transform large amounts of data and build complex, scalable data pipelines while saving time on backend confusion,” said Google.
“Serverless” is a misnomer; there are still servers in serverless computing. What’s missing is the management of the server hardware itself. Serverless computing simply abstracts the operations of the hardware needed to make cloud apps work.
Serverless offerings from hyperscalers include AWS Lambda, Microsoft Azure Functions, Google Cloud Functions and IB Cloud Code Engine.
“A cloud provider handles the routine work of provisioning, maintaining, and scaling the server infrastructure. Developers can simply package their code in containers for deployment,” explained Red Hat.
While not ideal for all cloud computing scenarios, serverless computing operations convey a number of advantages. Serverless operations can result in lower cost and complexity. Services are offered using a pay-as-you-go subscription model, and they make it easy for developers to rapidly deploy and update apps, with provisioning and scaling server and network operations all managed automatically.
Serverless computing also comes with attendant challenges. Performance and workload management can be affected by scaling. Monitoring and debugging becomes more challenging. There are also potential data security issues.
“Because companies are not assigned their own discrete physical servers, serverless providers will often be running code from several of their customers on a single server at any given time. This issue of sharing machinery with other parties is known as ‘multitenancy’ — think of several companies trying to lease and work in a single office at the same time. Multitenancy can affect application performance and, if the multi-tenant servers are not configured properly, could result in data exposure,” said Cloudflare.