KeenTune SIG is an operating system full-stack intelligent optimization product driven by AI algorithms and expert knowledge bases. environment for optimum performance. The product has achieved good tuning results in multiple business products, and the average tuning effect of several typical applications of physical machines, virtual machines, and containers is above 10%. Help users to obtain higher network throughput, lower network delay, more efficient service processing capability and faster service response under the same resource environment. At the same time, users can reduce the demand for hardware resources according to their own business needs, thereby reducing business operation costs.
Author: Hu Yuxi, head of KeenTune SIG.
Welcome more developers to join KeenTune SIG:
Website : https://openanolis.cn/sig/KeenTune
Mailing list: [email protected]
SIG target
1. For mainstream open source Linux, it provides a full-stack expert knowledge solidification platform and intelligent performance tuning tools.
2. Build a visual algorithm development platform, especially in the direction of parameter interpretability and parameter optimization to form a competitive algorithm framework.
Current Progress
1. Establish a SIG in OpenAnolis for long-term maintenance support.
2. The installation package is provided in the official YUM of Alinux and Anolis.
3. Released at 2021.10.21 Yunqi Conference, open source in codeup and github.
Introducing KeenTune
KeenTune is a product designed to provide intelligent full-stack performance tuning for Linux. This product provides OS with the ability to dynamically and statically synergize tuning through the effective combination of AI algorithms and expert knowledge bases.
KeenTune, as a Linux full-stack tuning product, has the following features:
1. Distributed architecture, keentuned, target, bench, and brain components can be deployed on demand, can be divided and combined, and can be installed and used, which can be effectively applied to single-machine and cluster scenarios; lightweight component deployment has little impact on business and can be quickly Realize the docking with the original product management and control side.
2. Dynamic and static collaborative tuning, which can effectively combine expert knowledge base with interpretable parameter tuning. Dynamic customized tuning results can effectively solidify and facilitate unified cluster configuration. Static experience tuning can also be easily converted into dynamic tuning for customized tuning of specific scenarios. Dynamic and static coordination can effectively improve tuning capabilities and efficiency.
3. Cross-platform, can run on x86 and arm versions of mainstream OS such as CentOS, Ubuntu, Debian, AnolisOS, and Alinux.
4. High-efficiency interpretable algorithm, self-developed algorithm hybrid model, organically combining interpretable dimensionality reduction algorithm, high-dimensional Bayes, surrogate model and other algorithms, the algorithm efficiency of parameter tuning is more than 300% higher than that of Bayes . At the same time, the improvement of the algorithm in high fidelity effectively solves the problem of performance data fluctuation in real-world scenarios.
5. Visualized algorithm platform, the visualization of algorithm process can facilitate algorithm developers to analyze data such as convergence efficiency and LOSS feedback, and effectively improve the efficiency of algorithm development.
-- over--
About the Dragon Lizard Community SIG
The SIG is open and strives to make the deliverables part of the community release, governed by the core members of the group, and can communicate with members of the group through the mailing list. There are currently more than 20 SIGs in the dragon lizard community , including hardware compatibility testing SIG, commercial secret software stack, high-performance storage technology SIG, Java language and virtual machine SIG, Cloud Kernel, OceanBase SIG, etc.
SIG website : http://openanolis.cn/sig
About the Dragon Lizard Community
OpenAnolis is a non-profit open source community composed of enterprises, institutions, institutions of higher learning, scientific research units, non-profit organizations, individuals , etc. on the basis of voluntary, equality, open source, and collaboration. The Dragon Lizard Community was established in September 2020 to build an open source, neutral and open Linux upstream distribution community and innovation platform.
The short-term goal is to develop Anolis OS as an alternative to CentOS and rebuild a distribution compatible with mainstream international Linux manufacturers. The medium and long-term goal is to explore and build a future-oriented operating system, establish a unified open source operating system ecosystem, incubate innovative open source projects, and prosper the open source ecosystem.
Longli OS 8.4 has been released, supports x86_64 and ARM64 architectures, is fully compatible with Intel, Feiteng, Haiguang, Zhaoxin, Kunpeng chips,and provides full-stack national secret support.
Welcome to download:
https://openanolis.cn/download
Join us in building an open source operating system for the future!
https://openanolis.cn