The Right Choice
The telecom industry is one of the most rapidly evolving sectors in the world, and it demands innovation, scalability, security, and reliability. In this ever-changing landscape, the choice of operating system can mean the difference between success and failure, scalability or a walled garden that prevents it from growing beyond current constraints.
openEuler, a community-driven, open source enterprise Linux operating system, offers a stable and secure foundation for the telecom industry and other critical enterprises. By leveraging the capabilities of open source software and innovative technologies, telecom organizations can achieve cost reductions, improved performance, enhanced security, and faster innovation cycles. The open source nature of openEuler also facilitates collaboration and innovation, ensuring that the platform remains up-to-date and meets the ever-changing needs of the industry.
Designed for servers, edge devices, and IoT, openEuler is an ideal option for an industry that requires real-time and deterministic performance, carrier-grade high availability, scalability and multi-core optimization, support for Network Functions Virtualization (NFV), hardened security and isolation, lightweight embedded support, and open APIs and telemetry integration.
What Is openEuler?
openEuler is an open source, community-driven operating system that has gained significant traction in the telecom industry. Its modular architecture, GPU support, and advanced security features make it an attractive option for companies looking to upgrade or expand their infrastructure.
Initiated by Huawei and made open source in 2019, openEuler was donated to the OpenAtom Foundation in November 2021. It is developed collaboratively by a broad community and supports multiple processor architectures, including Arm and x86.
With a focus on enterprise readiness, scalability, and modern technologies, openEuler is well-equipped to handle the demands of large-scale telecom installations. One key factor in this is openEuler's modularity, which is achieved through scenario-specific separation, decoupled software packaging, and architectural layers that allow it to scale across vastly different environments like servers, cloud, edge, and embedded devices. The scenario-specific modules include:
- openEuler Server/Cloud: This module is geared toward enterprise data centers with optimizations for containerization, memory management, and virtualization infrastructure.
- openEuler Edge: This lightweight module is stripped down to optimize performance on smaller edge computing gateway devices that are resource-limited.
- openEuler Embedded: This heavily modularized version of openEuler is built specifically for IoT devices and industrial controls. The embedded version features an elastic virtualization base that allows for the isolation and running of multiple operating systems concurrently on a single system on a chip (SoC).
Why Use openEuler in Telecom?
Like most distributions, openEuler benefits from the performance and modularity of the Linux kernel. But it's much more than that, as openEuler offers plenty of optimizations to the standard kernel (Figure 1), such as:
- Tidal Affinity Scheduling (TAS): Dynamically adjusts CPU affinity based on application load. Under light loads, TAS consolidates tasks onto preferred CPU cores to maximize resource locality and cache. Under heavy loads, TAS expands to new cores to preserve Quality of Service (QoS).
- Simultaneous Multithreading (SMT) Expeller without Priority Inversion: In SMT setups, this optimization avoids priority inversion to prevent low-priority or offline tasks from degrading the performance of mission-critical, online services.
- Dynamic Kernel Preemption: Adds dynamic runtime configuration (
preempt=none/voluntary/full) to ensure that the kernel can alter its preemption behaviors on the fly based on whether the system needs raw throughput or microsecond-level latency responsiveness.
The telecom industry has very specific requirements (see Figure 2), such as:
- Real-time and deterministic performance: Operating systems or real-time extensions (Real-Time Operating Systems) must handle packet routing, switching, and signaling with strict, predictable timing to prevent jitter (the variation in packet arrival time as data travels across a network) and dropped calls or data packets.
- Carrier-grade high availability (HA): Features like hitless software upgrades, automatic failover, process isolation, and self-healing kernels ensure a carrier's network stays active 24/7/365 without necessary intervention.
- Scalability and multicore Optimization: The OS must efficiently distribute heavy network traffic across multicore processors, multithreaded environments, and distributed cloud or edge nodes, in order to meet incredibly high demands.
- Support for NFV and Software-Defined Networking (SDN): Required to spin up virtual routers, firewalls, and core network functions dynamically.
- Hardened security and isolation: The telecom infrastructure handles critical national and enterprise communications, which means the OS requires strict kernel hardening, secure boot mechanisms, cryptographic acceleration, and strong process sandboxing combined to resist cyberattacks.
- Lightweight embedded support (for edge/base stations): The telecom industry relies heavily on cell towers, routers, and IoT gateways, so operating systems must maintain a very small footprint while at the same time supporting complex networking stacks.
- Open APIs and telemetry integration: The telecom industry depends on an industry-standard, programmable interface that allows operational support systems to gather real-time hardware metrics, traffic logs, and fault alerts.
openEuler at Work
Several telecom companies have leverages openEuler (Figure 3). The following case studies illustrate how openEuler has helped these organizations within the telecom industry succeed:
- China Telecom needed an enterprise-grade, highly secure, and cloud-native operating system to power its rapid digital transformation and massive cloud business. To achieve this, China Telecom developed its own customized operating system, CTyunOS, which was built directly on the openEuler foundation. In order to leverage openEuler’s migration tool systems, China Telecom seamlessly transitioned infrastructure, platform, and application layers to the new OS and was able to achieve over 30,000 CTyunOS instances across Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) nodes. Once completed, they achieved a performance boost of over 10 percent.
- China Mobile (Zhejiang) is a top provincial mobile carrier that handles complex user traffic. It needed to migrate its core support systems without causing downtime or affecting the user experience. To do this, they adopted Galaxy Kylin Advanced Server Operating System V10, which was derived from openEuler. The architecture was built on dual-plane x86 and Huawei Kunpeng server clusters running Mesos and Kubernetes (K8s) container clouds. This deployment endured high-traffic events such as a campus marketing rush to provide operational high availability and stability.
- China Mobile (Independent Core IT Infrastructure) looked to eliminate its dependency on legacy operating systems that were too restrictive by transitioning to open source technology. To do this, they integrated UnionTech OS Server 20, which is a commercial OS that is based on openEuler, with their existing container networks. This migration was the first instance of a major telecom provider successfully validating an independent compute stack on core telecommunication traffic and proved that large-scale architectures, focused on the telecom industry, could achieve full replacement without requiring costly app rebuilding.
- 5G and edge cloud ecosystems (e.g., Chuangpuyun) leverage openEuler for multi-concurrency edge computing that is required by 5G application infrastructure. These cloud providers serve the telecom edge space and use openEuler as their foundation, utilizing innovations such as openEuler's iSulad container engines, automated tuning toolkits (such as A-Tune), and KubEdge for cloud-edge collaborations. These deployments provide ultra-low latency frameworks that are required for graph computing, AI-assisted operations, and industrial Internet integrations.
So, what makes openEuler so well adapted for the telecom industry? First of all, openEuler is designed to handle high-performance computing (HPC) and large-scale deployments, which is perfectly in line with the needs off telecom companies that require fast and efficient processing. openEuler also includes proactive security hardening, regular updates, and patches to ensure that telecom companies are able to maintain the highest level of security for both their networks and the data that travels on them. As well, openEuler meets the stringent financial regulations required by the telecom industry, such as compliance with China's financial industry standards. Finally, openEuler supports modern features like containers, K8s, and AI frameworks, making it easy to integrate emerging technologies and stay ahead of the competition.
Unique Components of openEuler
openEuler offers scenario-based versions, such as openEuler Server, openEuler Edge, and openEuler Embedded; each of these versions is a viable option for the telecom industry. On top of that, openEuler uses a flexible virtualization base that combines bare metal, lightweight containers, LibOS, and trusted execution environments.
LibOS, short for Library Operating System, is an OS that is implemented as a collection of libraries linked directly to an application's address space. LibOS allows a program to interact with hardware and/or resources without having to rely on a standalone kernel. LibOS uses only a single address space and uses a unikernel to deliver high performance. Think of LibOS as an entire operating system, implemented as a library, that is ideal for cloud-native applications running in virtualized environments for efficient, low-overhead, and ultra-portable apps.
openEuler also depends on a modular container and runtime stack, which is achieved with tools like the iSulad container engine, the StratoVirt virtualization platform, and modular AI container images that can be swapped and/or scaled independently.
Finally, openEuler features the MICA Deployment Framework, a unified deployment layer that allows a multicore system to run a traditional Linux OS (such as Ubuntu) side-by-side with a Real-Time Operating System (such as Zephyr).
All of this comes together to serve the telecom industry in ways other operating systems cannot.
Real-World Applications
openEuler is not just a theoretical solution; it's been successfully implemented in various telecom companies around the world. For example, China Mobile has deployed openEuler to optimize network performance, while other companies have used it to build high-performance data centers and processing platforms.
Other real-world applications of openEuler include:
- Huawei/Huawei Cloud: Huawei relies heavily on openEuler to power its cloud infrastructure, enterprise servers, and its broader hardware ecosystem.
- Baidu/Baidu AI Cloud: Baidu uses openEuler within its data centers and has collaborated with other organizations to integrate features such as Arm Confidential Compute Architecture (CCA) and AI-native optimized toolchains into the OS itself.
The Role of Community and Ecosystem Partners
A strong community and equally strong partners have contributed to the success of openEuler. Companies like SUSE, IBM, and Red Hat have collaborated with openEuler to drive feature development, security fixes, and industry support. Other collaborators include:
- KylinSoft (enterprise-grade virtualization and the UKUI desktop project)
- UnionTech (commercial OS editions based on openEuler)
- iSoft (infrastructure work for GTK4 and Wayland porting)
- TurboLinux (commercial operating systems based on openEuler)
- Loongson Technology (helped adapt openEuler to LoongArch architecture)
- The Institute of Software Chinese Academy of Sciences (commercial editions of openEuler)
Conclusion
openEuler's importance to the telecom industry is clear: By leveraging the capabilities of open source software and innovative technologies, telecom organizations can achieve cost reductions, improved performance, enhanced security, and faster innovation cycles. openEuler's modular architecture, GPU support, advanced security features, support for diverse hardware platforms, and integration withsupport for processor architectures such as Arm, x86, and LoongArch and other processor architectures also enhance the platform's ability to meet the demands of modern telecom systems. As the telecom sector continues to evolve, openEuler's role in supporting the industry's growth and development will be critical for its success.
This article was made possible by support from openEuler by OpenAtom Foundation through Linux New Media’s Topic Subsidy Program.
