Dragon Lizard Boosts Server OS Development

Advertisements

In the present age of artificial intelligence (AI), China's server operating system industry is exhibiting unprecedented vitality, signaling a significant transformation in the technological landscapeThe recent announcement by CentOS to discontinue its maintenance has accelerated the development of alternative server operating systems in China, providing immense market opportunities for innovation and growthAs AI emerges as a critical factor influencing future innovations, it is driving a complete restructuring of server operating system products and their associated ecosystems.

One prominent example of this invigorated industry is the Longxia Community, which has thrived over the past five yearsThe community has witnessed the installation of over 8 million copies of the Longxia operating system across various versions and has successfully brought together more than 1,000 partners from both upstream and downstream sectors of the industry

These partnerships have led to widespread deployment and utilization of the Longxia operating system in industries such as internet services, finance, and government sectors, marking the emergence of a viable and sustainable open-source server operating system sector.

As Zhang Dong, the vice-chairman of the Longxia Community, aptly puts it, "Moving quickly alone is good, but going far requires collaboration." The community has established a balanced ecology involving core enterprises at various levels, avoiding dominance by any single entityThis collaborative approach motivates all participants to contribute actively to each aspect of the ecosystem, paving the way for long-term and healthy development of the community itself.

The advent of AI technology signifies a comprehensive alteration of operating systems, particularly server types, which play a crucial bridging role in data centers

Server operating systems are increasingly integrating advanced AI technologies, compelling them to evolve into smarter systems capable of managing diverse computational requirements and collaborating effectively with cutting-edge hardware, such as GPUs and DPUsFurthermore, these systems must support the monumental influx of intelligent applications currently proliferating in various sectors.

This radical change stems from the accelerated pace of AI advancements, which are reshaping the landscape of data center hardware and operating systemsHowever, as noted by Zhang Dong, infrastructure expansion has its boundaries; a careful balance must be struck between scale, energy consumption, and costsConsequently, server operating systems have substantial room for innovation in executing tasks in a more cost-effective and space-efficient manner.

The ongoing integration of AI into server operating systems is critical, presenting a significant direction for development—creating AI-native server operating systems that significantly enhance operational efficiency and usability

The Longxia operating system exemplifies this trend with features such as the AI assistant "Copilot," designed to make the user experience more convenient and efficientYang Yong, the chair of the Longxia Community's technical committee, emphasizes their continuous exploration of leveraging AI capabilities to improve user engagement.

Moreover, AI technologies represented by large models fundamentally alter the infrastructural scale and its inherent complexityServer operating systems must comprehensively adapt to large-scale, multifaceted computational environments, effectively manage resources, and oversee operationsA recent report from Meta illustrates the reliability challenges faced in managing infrastructures when its Llama 3 model experiences hundreds of fault errors during a 45-day training run on a massive cluster.

With the emergence of AI clusters ranging from thousands to tens of thousands of nodes in China, the challenges for server operating systems are becoming increasingly significant

alefox

Zhang Dong argues that the number of processor cores and task threads managed by operating systems is growing exponentially, increasing the complexity of memory management tasks as AI applications demand more considerable memory resources.

Recently, the Longxia Community released its newest version, Anolis OS 23, which drastically enhances management and scheduling capabilities within infrastructures, including multi-platform compatibilityIt ensures comprehensive support for various processors, algorithm libraries, and AI frameworks, fostering effective training and inference for large models.

As we move deeper into the AI era, server operating systems are expected to take on an increasingly crucial role, acting as the brain within infrastructure domainsThey are tasked with maintaining data center reliability, infrastructure observability, and holistic optimization.

Unlike other software products, server operating systems exhibit a high degree of user stickiness; once deployed in an operational environment, they typically remain in continuous use

The discontinuation of CentOS has indeed profoundly impacted Chinese enterprises by underscoring the necessity for new server operating systems that appropriately balance the past, engage with current needs, and maintain a focus on future advancements.

"Balancing the past" entails ensuring that new server operating systems remain compatible with previous environments and applications, guaranteeing a smooth transition for ongoing business operations"Engaging with the present" involves addressing contemporary challenges and the evolving trends of AI technology to meet the new demands from usersLastly, "looking towards the future" encompasses the need for clarity in technological and product development paths, boosting user confidence in upcoming software offerings.

In light of the current state of domestic enterprises, the Longxia Community has introduced three pivotal initiatives: the "CentOS Replacement Plan," the "Anolis OS 23 Ecosystem Derivative Plan," and the "AI Application Promotion Plan."

The "CentOS Replacement Plan" is grounded in pragmatic approaches, ensuring strategic takeovers, migrations, and extended protections for businesses after the cessation of CentOS services

For instance, Inspur Information, a member of the Longxia Community, has collaboratively launched the "No Worries After Suspension" initiative, designed to help customers mitigate potential security risks associated with the shutdown of CentOS and providing a variety of transition tools and strategies for seamless migration.

As open-source operating systems evolve, they inevitably face challenges of kernel inconsistencies, numerous derivative versions, and compatibility, complicating future product development, user experiences, and strategic planningThe Longxia Community’s "Anolis OS 23 Ecosystem Derivative Plan" seeks to unify kernel technology, supply chains, and other community-contributed standards through collaborative efforts across the industry chain.

"The fragmentation issues within operating systems pose significant challenges—all new products or configurations necessitate extensive testing," Zhang Dong points out

With the support of numerous industry chain partners, the "Anolis OS 23 Ecosystem Derivative Plan" has already resulted in the release of 12 derivative versionsFor example, Inspur’s commercial derivative cloud platform, KeyarchOS, based on Anolis OS 23, has undergone comprehensive upgrades in performance, intelligence, and security, with targeted optimizations for AI-native application scenarios, ideally suiting modern demands for server operating systems.

Acknowledging the rising trend of AI applications, the Longxia Community has also launched the "AI Application Promotion Plan," establishing a developmental roadmap for AI native operating systemsThis initiative revolves around AI container images, intelligent operations through AIOps, and documentation construction for OS Copilot, continuously advancing the integration of AI attributes within the Longxia operating system.

In the realm of foundational software, there is no such thing as taking shortcuts; server operating systems are no exception

Although substantial progress has been made in China’s open-source operating system sector, it still lags compared to more mature international ecosystemsThe success of a server operating system, recognized as the pinnacle of foundational software, relies on technological and product efficacy as its baseline and ecological effectiveness as its ceilingThe levels of technology and products must remain consistently advanced, capable of adapting to demanding and dynamic business scenarios.

Additionally, the ecosystem must exhibit growth and sustainability, welcoming extensive and in-depth participation from partners throughout the industry chain to cultivate a thriving ecosystemThus, the evolution of a new server operating system from inception to robust growth is akin to a gradual accumulation of effort leading to successYang Yong believes that the inherent complexity of operating systems necessitates continuous iteration and collaboration of technology, products, and ecosystem partners, aiming for collective growth through combined efforts at every step.

To ensure healthy and sustainable development, the Longxia Community is adopting a new approach to open-source operating systems, characterized by open and neutral decision-making mechanisms that unite industry partners