Cloud Native Technology Base
In 2015, Transwarp launched TCOS, a cloud native operating system that combines big data technology and container technology to manage and orchestrate complex big data services, enabling the orchestration, scheduling, deployment, configuration and operation & maintenance management of stateful big data services on the platform. With a solid and advanced technological base and dozens of invention patents, TCOS is an important technological innovator and pioneer in the field.
Supports Isolation of Multiple Types of Resources
TCOS resource management framework implements comprehensive resource isolation, covering different types of resources such as CPU, GPU, memory, network, hard disk, etc. It supports separation of management network and business network, isolation of GPU memory and computing power, fine-grained authentication and authorization management.
Federated Cloud Architecture
TCOS provides distributed data cloud services through multi-cloud computing power to support unified management of infrastructure resources in multi-cloud scenarios, including providing node resources management for clusters, elastic scaling of node resources, and visualization of infrastructure resources. Also, TCOS supports the deployment of federated cloud products and federate applications in a primary-standby mode or multi-active mode, implements cross-cluster migration of cloud products, applications and data, so as to support long-term evolution of enterprise-level data application cloud platforms.
Heterogeneous Hybrid Deployment
TCOS provides heterogeneous technical capabilities such as heterogeneous storage, heterogeneous network, heterogeneous CPU and GPU scheduling, thereby supporting X86, ARM architecture CPU and GPU, Linux, Windows and domestic operating systems, as well as Hybrid deployment on heterogeneous hardware and heterogeneous operations.
Advanced Orchestration and Scheduling Technology
TCOS provides multi-dimensional and scalable scheduling technology for stateful services, including local storage priority and uniform shceduling, local storage mutual exclusion scheduling, GPU exclusive/shared scheduling, network scheduling, node resource-aware scheduling, application resource-aware scheduling, resource reservation scheduling, and many other scheduling policies.
Unified Storage Access and Management
TCOS can integrate and manage various types of storage systems, and is suitable for systems requiring high IO performance and high availability. It is designed to better meet the diverse needs of storage systems for differentiated data services in data lake scenarios. TCOS also provides standardized access to multiple storage plug-ins and a unified storage monitoring and operation & maintenance platform to achieve management and operation & maintenance of a variety of backend storage systems.
Comprehensive Security Protection
TCOS implements private computing security zone based on zero-trust security technology to support the demand for computing security, network security, and data security when big data services perform federated learning and trusted computing. Through hardware, access encryption, storage encryption, computational process encryption, east-west traffic control, north-south traffic control, whitelist and other security technologies, it solves the difficulty of data protection and application integration, and supports the construction of data circulation ecosystem.