TKE supports dynamic scaling of clusters and vertical scaling of nodes.
The nodes in the cluster are manageable and deployable across availability zones, and the containers in the service are schedulable across availability zones.
The nodes in the cluster are provided with
TKE supports dynamic scaling of clusters and vertical scaling of nodes.
The nodes in the cluster are manageable and deployable across availability zones, and the containers in the service are schedulable across availability zones.
The nodes in the cluster are provided with
Improved service deployment efficiency
Service version is controlled through templates, and the image guarantees environmental consistency, allowing for faster service migration.
Service discovery is supported. Services can be accessed through load balancing domain names or service names together with port number, circumventing the impact of IP changes when the service backend changes.
Micro-services are supported, reducing
Improved service deployment efficiency
Service version is controlled through templates, and the image guarantees environmental consistency, allowing for faster service migration.
Service discovery is supported. Services can be accessed through load balancing domain names or service names together with port number, circumventing the impact of IP changes when the service backend changes.
Micro-services are supported, reducing
Manageable business configurations of different environments
Different environments for the same applications can be deployed, making it easy to update and roll back the applications.
Multiple versions are supported, but only
Manageable business configurations of different environments
Different environments for the same applications can be deployed, making it easy to update and roll back the applications.
Multiple versions are supported, but only
Official Dockerhub image management
Accelerated retrieval of official Dockerhub images is available.
Private image management
A secure and reliable private image repository is provided.
Fast image upload and download is
Official Dockerhub image management
Accelerated retrieval of official Dockerhub images is available.
Private image management
A secure and reliable private image repository is provided.
Fast image upload and download is
TKE supports dynamic scaling of clusters and vertical scaling of nodes.
The nodes in the cluster are manageable and deployable across availability zones, and the containers in the service are schedulable across availability zones.
The nodes in the cluster are provided with
Improved service deployment efficiency
Service version is controlled through templates, and the image guarantees environmental consistency, allowing for faster service migration.
Service discovery is supported. Services can be accessed through load balancing domain names or service names together with port number, circumventing the impact of IP changes when the service backend changes.
Micro-services are supported, reducing
Manageable business configurations of different environments
Different environments for the same applications can be deployed, making it easy to update and roll back the applications.
Multiple versions are supported, but only
Official Dockerhub image management
Accelerated retrieval of official Dockerhub images is available.
Private image management
A secure and reliable private image repository is provided.
Fast image upload and download is
- Security isolation: Each sandbox runs in an isolated controlled environment.
- Ultimate startup speed: Instances launch in milliseconds, ensuring Intelligent Agent calls are out-of-the-box.
- Various types: built-in browser sandbox, code sandbox, support scalable custom sandbox.
- Multiple access methods: sandbox interface and protocols compatible with mainstream open-source communities
- Inference acceleration: Tencent Self-developed TACO inference acceleration framework for all-round improvement in inference performance.
- GPU sharing: Deeply integrated qGPU sharing technology to fully leverage heterogeneous computing power
- Ultimate elasticity: super node pre-created sandbox technology, combined with HPA/VPA to achieve millisecond-level response for reasoning resources
- High-speed interconnection: Supports high-performance networks such as RDMA to ensure low-delay data transmission.
- Professional scheduling and fault tolerance: Optimizing resource allocation and topology awareness for distributed jobs
- Self-healing: When a node fault occurs, the training task can automatically Checkpoint and restore based on fault notification, ensuring training duration and achievement.
- Workflow orchestration and scheduling: Supports mainstream cloud native workflow engines such as Argo Workflows, enabling easy orchestration of complex data preprocessing pipelines through containerized tasks.
- Efficient storage mounting: Seamless integration of high-performance cloud storage (such as CFS Turbo, Goosefs) enables fast, large-scale container data access through CSI plugin implementation.
- Cost optimization: Support offline hybrid deployment to significantly reduce data processing costs.
- Agile development and delivery: Provide standardized delivery carriers, combine GitOps/CI/CD pipelines, quickly build, test, and deploy applications, and accelerate service iteration cycles.
- End-to-end observability: Unified logs, monitoring, and alarm platform covering every level from applications, containers to the cluster.
- Stable and reliable: high availability deployment with backup and recovery support
Frequently Asked
Questions
You can go to the TKE high-risk operation to view and understand.
To effectively use Tencent Cloud container service (Tencent Kubernetes Engine (TKE)), you can check the quick start document for deploying TKE.
When using TKE, you may encounter multiple scenarios that require service authorization to access related Tencent Cloud resources. Each scenario usually corresponds to preset policies contained in different roles, mainly involving two roles: TKE_QCSRole and IPAMDofTKE_QCSRole. Go to the service authorization related role permission description for more info.
Frequently Asked
Questions
You can go to the TKE high-risk operation to view and understand.
To effectively use Tencent Cloud container service (Tencent Kubernetes Engine (TKE)), you can check the quick start document for deploying TKE.
When using TKE, you may encounter multiple scenarios that require service authorization to access related Tencent Cloud resources. Each scenario usually corresponds to preset policies contained in different roles, mainly involving two roles: TKE_QCSRole and IPAMDofTKE_QCSRole. Go to the service authorization related role permission description for more info.