Welcome to 'Introduction to Serverless Computing.' After watching this video, you will be able to: Define 'serverless computing'
and describe its concepts, and identify the key characteristics
of serverless computing. The Cloud Native Computing Foundation
(or CNCF) defines serverless as "The concept of building and running applications
that do not require server management. "It describes a finer-grained
deployment model where applications, bundled as one or more functions, are uploaded to a platform and then executed, scaled, and billed in response to the
exact demand needed at the moment." In other words, serverless computing offloads
the responsibility of infrastructure management to cloud providers, enabling developers to
focus on the application business logic. Think of serverless computing as a combination of function-as-a-service (or FaaS) platforms and backend-as-a-service (or BaaS) services. FaaS platforms are used to run functions. BaaS represents backend cloud
services, such as cloud databases, object storage services, and message queues. Now, the history of IT computing shows a progressive
trend from traditional computing to serverless computing, leading to faster deployments,
shorter lifespans, and increased productivity. Several milestones mark this evolutional trend: Traditional computing uses physical
machines but is hampered by upfront investments, capacity planning, and more. Physical machines take weeks or months to
deploy and have a lifespan of several years. Virtualization technology is implemented
in cloud computing for faster provisioning, high scalability and availability. Virtualization facilitates the creation of
multiple virtual machines (VMs) or containers. And VMs take minutes to deploy and
have a lifespan of days or weeks. Containers, which are based on
operating system virtualization, require seconds to deploy and
live for minutes or hours. And, serverless applications require only
serverless architecture for the core code. Serverless functions take milliseconds to
deploy and have a lifespan of seconds. Serverless computing is an advancement of
standard cloud computing, which abstracts both infrastructure and software environments. It is an architectural style where
code runs using a cloud platform, and the cloud provider manages the hardware and
software setup, security, scalability, and so on. The client is billed only for
usage and not for CPU idle time. And developers only need to focus on their
application code in the form of functions. Serverless computing exhibits
the following characteristics: It is hostless, meaning developers do not have
to procure, manage, and maintain servers. It’s elastic because autoscaling is
immediate and inherent for serverless. It offers automated load balancing that distributes the incoming traffic
across multiple backend systems. It is stateless, which results in faster
performance and higher scalability. It’s event driven, meaning functions
are triggered only when events occur. It provides high availability
with no extra effort or cost. And it is usage-based with granular billing. So, how do functions work in a serverless world? Consider, for example, the steps required
to create a Docker or Kubernetes container. The developer creates a function by writing code
in a language supported by the cloud provider, such as Python, Java, Node.js, C#, or Go. Next, the developer uploads
a function to the cloud. Then, events are defined that trigger
the function, such as a user click. And, once the event occurs, the trigger is invoked,
and the cloud provider runs the function, resulting in the container object. Having developers spend more time creating
high-quality and optimized applications greatly benefits organizations. Developers can build functions using any
popular programming language, extend the functionality of apps
by adding additional features, perform better testing since functions
perform only one task at a time, optimize apps and functions, and improve the user experience. To accomplish the goal of zero
operational considerations, cloud providers take responsibility for
routine infrastructure management and maintenance tasks such as maximizing compute, memory,
and networking utilization while minimizing costs, providing server management that
includes OS updates and security patches, enabling autoscaling, maintaining high availability, implementing security, configuring high performance (or low latency), and setting up monitoring and logging. Now in comparing different cloud service models
shown in this figure, let’s look at your responsibilities versus the cloud provider,
and more so, regarding the serverless service model. We start with the extreme left stack
(traditional), which you manage entirely. In the IaaS model, you manage the
layers from the OS up to the top, while the cloud provider
manages the lower four layers. In the PaaS model, you manage the
first two layers, Application and Data, while the cloud provider
manages the remaining layers. In the Serverless model, you
manage only the Application layer, while the cloud provider
manages the remaining layers. And in the SaaS model, the cloud
provider manages the entire stack. In this video, you learned that: Serverless is the concept of building and running applications that do not
require server management, serverless computing uses BaaS
services and FaaS platforms, the history of IT computing shows
an increasing productivity trend from traditional to serverless computing, with serverless computing, cloud providers
take care of infrastructure management, such as server management, OS updates,
security patching, and monitoring, serverless computing enables developers
to focus on applications and code, and finally, users are billed only for
usage and not for CPU idle time.