Container Orchestration

Glossary

"Container Orchestration"

This glossary explains various keywords that will help you understand the mindset necessary for data utilization and successful DX.
This time, let's take a look at "container orchestration," a technology that is becoming widely used in modern software development, especially in software development on the cloud.

What is container orchestration?

Container orchestration is an automation technology that manages, operates, and controls a large number of containers in system development and operation using container technology.
Recently, microservices have been increasingly adopted as an IT system architecture, and container orchestration has become important as a fundamental technology for realizing microservices using container technology.

The advent of container technology has ushered in an era in which "massive containers" are being used

The premise of container orchestration is the existence of container technology, which has become widely used. This technology allows for the separation of software execution environments, similar to virtual machines, to be achieved efficiently with little overhead. For more information, see below.

⇒Container | Glossary

Because container technology has made it easy to separate execution environments, containers will no longer be used only out of a need to separate execution environments, but will be used more proactively.

Unlike costly virtual machines, containers were easy to use and increasingly available. This led to the widespread use of large numbers of containers. Microservices, a system architecture gaining attention in the cloud era, involves system development based on the idea of composing a system from many small, independent parts (services), and these many small parts are realized using many containers using container technology.

⇒Microservices | Glossary

Anyway, with the advent of this convenient technology called container technology, containers are increasingly being used in a variety of ways, such as microservices, which are realized using a large number of containers.

The problem of what to do with the large number of containers

If you are doing something similar to what was done with virtual machines, such as separating the execution environments of an accounting system and a production management system by placing each system in a container, then humans will be able to manage and operate the containers.

However, with a microservices architecture, for example, an accounting system might be made up of dozens of containers, with additional containers being launched depending on the situation. When it comes to cloud services built with microservices, services might be made up of groups of hundreds or even thousands of containers. In such cases, it becomes impossible to manage them manually, and an operational platform that can automatically operate large numbers of containers becomes necessary. This is container orchestration technology.

For example, it is necessary to know which machine each of a large number of containers is running on, as well as the execution status and execution logs of each container. This state in which the situation is "visible" is said to have observability, and an operational infrastructure that can maintain high observability in an environment with a large number of containers is required.

Furthermore, just like the health monitoring used in traditional operations, it is necessary to make the system resilient by monitoring whether the software inside the container is running properly, and if there is a problem, automatically taking predetermined actions such as stopping and restarting the container to maintain system functionality.It is also necessary to perform so-called auto-scaling, which automatically starts a large number of specified containers when the processing load is high and automatically reduces the number of containers when the load subsides, depending on the processing load situation.

Or perhaps it is only with an automated container operation platform that has such functionality that developing and operating systems using microservices becomes a realistic possibility.

How to automate large-scale container management

In traditional IT infrastructure management, infrastructure engineers often monitored the status of servers and performed necessary operations manually. With container orchestration technology, container operations are basically automated, and how to operate a group of containers is determined by writing configuration files, similar to programming.

Another characteristic is that rather than describing what to do, you often write the desired container state (declarative description) and then have container orchestration technology realize that state.

For example, you can describe the state you want to achieve, such as distributing this type of container across multiple servers, keeping at least three running, monitoring their status and discarding the container if there is no response, automatically increasing the number of containers if the average processing load per container exceeds 70%, and automatically decreasing the number of containers if it is below 30%, and have the system operate automatically according to that state.

Previously, operations were performed manually by humans, but now we are living in a world where programming allows us to automate the operation of large numbers of containers, for example, over 10,000. This type of change is sometimes called Infrastructure as Code (IaC).

As a means of realizing the system architecture

In traditional system development, for example, you would analyze the requirements to determine what kind of system you needed to create, and then design the entire system based on that in an object-oriented manner. For example, each class and instance would be a component of the system, and these would be assembled into an overall design like classes or modules.

Microservices involve breaking down the functions of an IT system into smaller units and placing each unit in a container as an independent service. The components of the system become containers, and the means of assembling these components to assemble the entire system is called container orchestration technology.

It is a technology that automates operations and also functions as a technology that realizes the architecture of IT systems. This also means that the boundary between the operation of traditional IT infrastructure and the design of system architecture is disappearing.

A means to achieve cloud native (in the narrow sense)

Given this situation, "cloud native" is sometimes said to refer to a microservices architecture that uses containers. The following definition from the CNCF is often cited:

CNCF Cloud Native Definition v

Cloud-native technologies provide organizations with the ability to build and run scalable applications in modern, dynamic environments, including public, private, and hybrid clouds. Containers, service meshes, microservices, immutable infrastructure, and declarative APIs are some examples of this approach.

However, the above definition comes from the CNCF (Cloud Native Computing Foundation), which is working to popularize and promote container technology. There are many other things besides containers that have changed in thinking in the cloud era, and just because something does not use containers (or microservices) does not mean it is not a cloud-era technology.

Examples of how container orchestration technology is used

There are many cloud services that use microservices and container orchestration technologies, but it is well known that AWS and Netflix have fully adopted microservices from an early stage.

Netflix has adopted Kubernetes as its container orchestration technology, enabling it to implement extremely large-scale microservices using tens of thousands of containers. As its business grows, Uber is working to migrate its cloud services, which were originally run as monolithic applications (not microservices), to a microservices architecture.

Container orchestration technology also provided as a cloud service

There are several software programs available for automatically managing such container swarms, but the most well-known and widely used is "Kubernetes," developed by Google.

In addition, there is no need to set up your own container orchestration software and create an environment where you can use containers; each cloud provides a pre-set up, ready-to-use environment as a cloud service.

By using such a service, you can develop and prepare your own microservices in the form of containers, write and provide configuration files specifying how you want them to be operated, and the system will be automatically operated, allowing you to use the IT system you have created or provide it to an external party.

Each cloud offers a number of services related to containers, but just to name a few, AWS offers "Amazon ECS (Elastic Container Service)" and "Amazon EKS (Elastic Kubernetes Service)," Google Cloud offers "Google Kubernetes Engine," and Microsoft Azure offers "Azure Kubernetes Service."

Related keywords (for further understanding)

container
- Containers are a technology that creates an execution environment for multiple independent application systems within a single OS execution environment. Containers perform OS-level virtualization, as opposed to the hardware-level virtualization that has been used up until now.
Microservices
- Microservices is a new system development style that has become popular thanks to container technology. The concept is to proactively create system components as microservices, which are small, independent units, and to configure the entire system from a large number of microservices.
Cloud Native
- The development of cloud services or systems on cloud services is based on very different assumptions than traditional system development, which often involved developing software that runs on individual machines. This means that approaches suited to the cloud era are now required, using different ways of thinking, different methods, and different means.

HULFT 10: File Integration Middleware for the Cloud and Container Era

From the days of mainframes to the present day cloud computing era, HULFT has achieved seamless data integration across a variety of IT system development styles and technical environments.

HULFT 10, the de facto standard file sharing middleware in Japan

Please try out HULFT, the pinnacle of domestic MFT products with an overwhelming track record in Japan and the de facto standard for file integration platforms.

HULFT emerged during the technological transition from the mainframe era to the UNIX era as a highly practical technology that could link old and new technology environments, with extremely high reliability and was adopted by financial institutions' systems.It has since evolved into a means of data integration that absorbs the differences in all environments, supporting Windows and Linux, and now also supports seamless integration with cloud services.

File integration may seem like an old technology, but it is now being considered for use with cloud services, and the latest version, "HULFT 10," has been developed to enable smooth use in container environments.

While mainframes are still in use today, there are fields where technologies, ways of thinking, and priorities differ, such as Windows, Linux, integration with cloud-based object storage such as Amazon S3, and even the world of development using container technology. Utilize HULFT as a means for engineers in each field to effortlessly data integration with each other.

⇒ Learn about file transfer mechanisms through HULFT product introduction and online seminars
⇒ HULFT-WebConnect Product Introduction and Online Seminar
⇒ HULFT 10 for Container Services