HUAWEI CLOUD 5 Open Source Projects Released and Updated to Help Developers Achieve Application Innovation

Abstract: During the HUAWEI Developer Conference 2023 (Cloud), the sub-forum of "Release and update of 5 major open source projects, various underlying capabilities to help developers achieve application innovation" led by HUAWEI CLOUD Open Source ended successfully.

This article is shared from the HUAWEI CLOUD community " HDC.Cloud 2023 | HUAWEI CLOUD 5 Major Open Source Project Releases and Updates, Helping Developers Achieve Application Innovation ", author: HUAWEI CLOUD Open Source.

During the HUAWEI Developer Conference 2023 (Cloud), the sub-forum on "5 Major Open Source Projects Released and Updated, Various Underlying Capabilities Helping Developers Achieve Application Innovation" led by HUAWEI CLOUD Open Source ended successfully.

In this sub-forum, Deng Mingkun, general manager of HUAWEI CLOUD open source business, introduced the development status and plan of HUAWEI CLOUD open source, and officially released 2 open source projects: OpenTiny, Kmesh, and an interpretation of the feature upgrades of the 3 major open source projects.

Deng Mingkun, General Manager of Huawei Cloud Open Source Business

HUAWEI CLOUD officially releases 2 open source projects

New release project 1: Cross-terminal, cross-framework, cross-version enterprise application front-end component library—OpenTiny

OpenTiny is Huawei Cloud's open source web application front-end development kit, covering Vue2/Vue3/Angular multi-technology stacks, and has tool libraries such as theme configuration system/middle and background templates/CLI command line. After nine years of continuous polishing, OpenTiny has served thousands of projects inside and outside Huawei. It is a tried-and-tested, stable and reliable front-end development solution.

Core highlights:

1. Support cross-device and cross-framework: adopt Renderless non-rendering component design architecture, separate component logic into framework-independent pure functions, add an adaptation layer for different frameworks, and realize cross-device, cross-framework, and cross-version.

2. Rich scenarios: After 9 years of continuous polishing of 150+ enterprise-level components, serving more than 1,500 businesses within Huawei, meeting the ultimate requirements of Huawei's internal IT system for functions, performance, and reliability, it is stable, reliable, and safe.

3. Support configurable components: It can support the usage of both tagged components and configurable components, which is suitable for use in low-code platforms without secondary packaging.

4. Complete ecology: not only provides component libraries of different frameworks, but also provides a series of development tools, including CLI scaffolding tools covering the entire process of front-end development, out-of-the-box middle and back-end templates, and a theme configuration system that improves efficiency.

New Release Project 2: High Performance Service Grid—Kmesh

As a cloud-native next-generation technology, the service grid has gradually become the standard configuration of cloud infrastructure, but its sidecar architecture has performance problems, which is the key to the current grid technology promotion; Kmesh high-performance service grid release, through architecture innovation for Developers bring a new grid performance experience and realize the native service grid data surface capability of OS. Based on ebpf+ programmable kernel technology, the traffic management is lowered to the OS, and the access performance of grid services is greatly improved; core highlights:

sidecarless: There is no need to deploy proxy components on the data plane, the traffic management sinks to the OS, and the traffic orchestration work is completed along with the flow;
High performance: The service forwarding delay in the grid is reduced by 60%+ compared with the industry solution;
Security isolation: implement safe and reliable traffic orchestration based on ebpf, and achieve cgroup-level orchestration isolation;
Open ecology: support XDS protocol standard, automatic docking with Isito and other software;

HUAWEI CLOUD Three major open source project feature upgrades

openGemini: High-performance log data retrieval, helping to solve the problem of high cardinality of time series

The goal of the time-series database openGemini is to create a dedicated operation and maintenance engine that takes into account both massive data storage and analysis and retrieval. This time it released two key features in the field of operation and maintenance: log retrieval and high-cardinality storage engine.

On the one hand, the log is a special kind of time series data, and it is also very important system data in the field of operation and maintenance. The log retrieval feature released this time is mainly to solve the high storage cost, performance and stability brought by the storage of massive log data to the system. question.

On the other hand, there are many high-cardinality problem scenarios in the field of operation and maintenance. For this reason, openGemini has developed a high-cardinality storage engine, which uses more appropriate data clustering and sorting methods, and improves data filtering by building cardinality-independent sparse indexes. Effect and query performance, to fundamentally solve the high cardinality problem that has plagued time series databases for a long time.

Spring Cloud Huawei: Service Governance Upgrade, Helping Developers Realize Highly Available Cloud Native Microservices

The main goal of SpringCloud Huawei is to make it easier and more efficient for developers to use Spring Cloud to develop microservice applications and seamlessly integrate with Huawei Cloud services. In the past few years, Spring Cloud Huawei has insisted on giving back to the open source ecology, is not bound to Huawei cloud services, is compatible with Spring Cloud development habits and extensions, and provides developers with one-stop out-of-the-box registration discovery, centralized configuration, and service governance capabilities .
Feature upgrade:

Support request context. A unified request context can be used in technologies such as Feign, RestTemplate, WebMvc, and WebFlux to reduce the difficulty and quantity of code development for functions such as authentication, call chain, and gray release.
Support client isolation warehouse and client fuse. Using a simple stand-alone algorithm, it provides dynamic, distributed fault detection and overload protection capabilities.
Supports tag-based traffic control policies to meet the needs of reliability scenarios such as traffic peak shaving and valley removal, and fair use of resources.
Spring Cloud Huawei has been widely used in government, banking, e-commerce, and traditional manufacturing industries.

Sermant: A major upgrade of the cloud-native agentless service grid, bringing a new experience to non-intrusive service governance

As a cloud-native agentless service grid, Sermant is committed to solving the rapid evolution of microservice governance functions in the cloud-native era from three technical directions: non-intrusive, high-performance, and plug-in. Compared with the SDK, business applications are connected to Sermant in a non-intrusive way with zero code transformation, and the performance of Sermant after access is significantly improved compared with Service Mesh. Excellent experience.

Feature upgrade:

Agent observability improvement: Without any additional auxiliary middleware, you can know the health status of all Sermant Agents, including any abnormal conditions in the process, and through the built-in monitoring service, it is clear on one screen.

2. Model unification and link coloring based on routing rules: Based on unified rules, it supports traffic governance in various complex scenarios, such as full-link grayscale release, pressure test traffic Mock, and call routing in the same computer room.

3. System rule flow control and load self-adaptive flow control: current limiting and downgrading bid farewell to each application manually filling in parameters, and the system dynamically performs intelligent flow control according to its own situation.

4. Elimination of outlier instances: eliminate avalanche failures of microservices caused by single instance exceptions.

5. Enhanced availability management capabilities: avoid the unavailability of services in the entire computer room due to a single service failure in the same computer room in scenarios such as routing in the same computer room.

During the event, the agenda of the HUAWEI CLOUD open source sub-forum attracted many participants to the venue. In addition, HUAWEI CLOUD Open Source also deeply participated in the exhibition area, CodeLabs training camp, geek challenge and other links, and conducted in-depth exchanges and interactions with developers on the spot. In the future, HUAWEI CLOUD Open Source will continue to adhere to open source and openness, practice technology inclusiveness, and enable developers to easily use cloud-native technologies and quickly build cloud-native applications.

Click to follow and learn about Huawei Cloud's fresh technologies for the first time~