How TCOS 2.0 Empowers Big Data, AI, and Cloud Workloads with Enhanced Compatibility
TCOS 2.0, the container operating system from Transwarp, expands compatibility to Windows, ARM, MIPS, and domestic platforms, adds GPU heterogeneous scheduling, HPA autoscaling, enhanced local storage management, and improved monitoring, providing a robust foundation for big data, AI, and cloud-native applications.
TCOS (Transwarp Container Operating System) is a container cloud OS designed for big data, AI and cloud computing, providing underlying support for containerized workloads.
Version 2.0 Release
Released on 2020-04-10, TCOS 2.0 is compatible with Kubernetes 1.16 and Docker 17.12, synchronizing with the open‑source community while extending platform capabilities.
Compatibility Enhancements
TCOS 2.0 adds support for Windows Server 2019, enabling heterogeneous Windows + Linux clusters, and expands compatibility to domestic ARM (Huawei Kunpeng 920, Phytium 2000+), MIPS (Loongson‑3), and Chinese operating systems such as Kylin and NeoKylin.
Deployment Service Optimization
The Web Installer 2.0 offers a smoother installation experience, automatic license reading, and information query, and supports seamless upgrades from 1.9/1.10 to 2.0.
HPA Autoscaling Support
TCOS 2.0 integrates the Kubernetes Horizontal Pod Autoscaler (HPA) to automatically scale pods based on CPU, memory, and other resource metrics.
GPU Heterogeneous Support and Monitoring
The platform provides heterogeneous GPU scheduling, allowing users to target specific GPU types, and introduces a GPU‑Monitoring‑Tool that tracks temperature, memory usage, and utilization.
Local Storage Optimization
TCOS 2.0 enhances local storage with multiple pool types (HDD, SSD, RAID 1, RAID 5, raw disks), supports PVC anti‑affinity rules to avoid I/O contention, and adds a data‑migration tool (Transwarp‑Pod‑Migration‑Operator) for stateful services.
Monitoring and Alert Improvements
Node‑Problem‑Detector monitors kernel and plugin health, reporting node status, while integrated health checks cover network and storage components, offering cluster‑ and tenant‑level resource overviews.
Future Directions
Building a domestic ecosystem with broader hardware compatibility.
Supporting heterogeneous federated deployments across data centers.
Dynamic resource scheduling, pod in‑place updates, and enhanced container security.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
StarRing Big Data Open Lab
Focused on big data technology research, exploring the Big Data era | [email protected]
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.