Shulou(Shulou.com)06/02 Report--

In the process of refactoring and updating applications, there are often some challenges. The more often you update your application, the more complexity you will increase. Getting applications to run on the container platform, and they can communicate and connect with each other, is the only way to a modular and flexible micro-service architecture. But this flexibility of microservices also makes it more complex. Then it's Service Mesh's turn to play a role!

Service Mesh provides enterprises with the centralized control panels they need while still being able to use flexible, cloud-based application development. We can think of Service Mesh as a dedicated layer 7 grid for micro-service API, which provides authentication, authorization, security, and performance services to optimize "east/west" traffic between services. More importantly, it provides you with a central point to apply to these strategies without having to code all of them directly into the application's business logic.

Simple Service Mesh analogy

Service Mesh is like the plumbing network of a city. Your team controls these pipes, connects them as needed and sets up all flow controls between them. No matter what type or purpose, or the requirements of the applications supported by Service Mesh are constantly changing, data can be passed through your system.

This flow control can be done at a central location, where rules are built to manage interconnected data flows. It's like a giant control room in the sky, where you can water the land in California when crops need extra resources, or Miami is so wet that you can drain them. The most important point is that these operations can be performed automatically and adjusted dynamically.

Service Mesh enhances reliability and visualization

Service Mesh provides intelligent traffic routing that automatically recovers from network or service failures, so that problems throughout the stack can be traced, even interruption between services.

If the server does not respond, your service grid will remove it from a single service, or from an active, load-balanced service pool, and move it to another pool, which often checks to see if it is operational. When the server starts to respond within a reasonable time range, it is automatically push back into the active load balancing pool.

By providing visualization of all aspects of the service layer system, Service Mesh can also be used to debug and optimize the system. In this way, the dirty water problem (murky water) in microservices is solved. Over time, the system can be adjusted to expand functionality to meet performance and stability requirements.

Service Mesh protect inter-service communication

When your team releases a new version of the application, or when you want to migrate the application-hosted cluster to a new data center, the security team usually needs to reissue the certificate and authorize it to the new server in the system. This will take a lot of time and energy and is an obstacle to promoting production improvement.

With the service grid, the security of communication between services is handled by the grid. These concerns are abstracted from the application itself, and all these limitations are handled by the service grid, such as which services can communicate with each other, which systems can access which services, and which users can access which services. Therefore, upgrading applications in the grid does not require reallocation of security resources.

This also allows security issues surrounding communication between networks and services to be separated from any internally developed business logic. If there is a security breach in the network build, the service grid handles changes around security updates rather than rearchitecting each application. This eliminates a large amount of downtime when making security changes and updates.

Research on Service Grid in large-scale Micro-Service Environment

However, the service grid has a (huge) potential drawback. It adds extra containers, and in fact, it doubles the size of the container. Most service grid implementations use sidecar proxies, coupling a proxy instance to the microservices bound by each container. In this way, the benefits outweigh the operating costs, which means that the service grid is often too large for small environments.

However, if you are managing dozens or even hundreds of separate micro-services, consider a service grid. With the service grid, your team can better track issues, ensure the availability of services, and maintain the correct distribution of routing tables. For these large environments, whether in the public cloud, in your enterprise data center, or in the implementation of a hybrid cloud, they are the last piece of the cloud application puzzle and a key part of connecting your entire industry.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.