Industry news
one Cloud native top technology event will be launched soon, and the keynote speech+project highlights are here!
Summary: From December 9 to 10, 2021, the "KubeCon+CloudNativeCon+Open Source Summit China 2021" jointly sponsored by CNCF and global open source contributors will meet with cloud native enthusiasts in the form of virtual conference. In this year's keynote speech, nearly 20 senior cloud native technology experts from CNCF and major manufacturers will bring wonderful sharing, including both the speech topics on CNCF and Kubernetes, as well as the technology exploration on financial cloud native, multi cluster management, security, cloud future and other leading industries.
Cloud Native Best Practices
one Use Service to implement simple grayscale publishing and blue-green publishing
Summary: The principle mainly involves that users usually use Kubernetes objects such as stateless load deployment and stateful load StatefulSet to deploy services, and each workload manages a group of pods. Usually, a corresponding service will be created for each workload. The service uses the selector to match the pod. Other services or services outside the cluster can access the services provided by the back-end pod by accessing the service.
two GRPC error handling practices
Summary: All service providers are required to define the protobuf file of the interface and error code under an independent path in advance, and then submit it to GitLab. The format, lint, and breaking of the changed protobuf file are checked through the check phase of GitLab CI. Then in the build phase, documents will be automatically generated based on the comments in the protobuf file and pushed to the internal microservice management system interface platform. Pile codes and error codes in Go/PHP/Node/Java and other languages will be automatically constructed based on the protobuf file and pushed to the designated centralized warehouse.
three Homework Cloud original cost optimization practice
Summary: With the development of business requirements, JobHelp's IT system is facing huge challenges. The existing infrastructure platform architecture has been unable to meet the rapidly growing business needs. After using HPA and offline hybrid deployment, JOB helps the business to rapidly iterate, rapidly expand and shrink the capacity in seconds, implement the service operation state specification and unified operation and maintenance environment, and unify the multi cloud environment to improve service availability.
Recommended reading
one It's 2021. What is the community discussing about service grid?
Summary: This article mentioned that from container orchestration to service grid, it can be seen that cloud computing is a standard struggle, even the whole cloud computing; The goal of service grid is to become a cloud native network infrastructure, which has a long way to go; Service grid is just one part of the huge cloud native technology stack. Don't block the vision of the whole world. The development of technology is endless; Istio architecture has been stable, production is available, and ecology is in the bud; For end users, what is suitable for them is the best.
two Kubernetes Deployment Troubleshooting Visual Guide (2021 Chinese Version)
Summary: Deployment Service、Ingress， The whole application goes through a long process from deployment to normal operation. From the kubectl apply YAML file, through apiserver controller manager、scheduler、kubelet、 As well as the collaborative work of CRI, CNI and many other components, this article gives a view on deployment troubleshooting.
three Service grid without Sidecar proxy based on gRPC and Istio
Summary: The experimental features supported in Istio 1.11 can directly add gRPC services to Istio without injecting Envoy proxy into Pod. This article also provides a Demo performance test data, which can greatly improve application performance and reduce network latency.
