Container Instance Service

A convenient, secure, affordable and flexible fully-managed container service that is compatible with Kubernetes

Overview

Container Instance Service (CIS) can help you quickly and flexibly deploy containers in the cloud, allowing you to focus on building programs and using containers instead of managing devices. With no pre-purchased CVMs required, a batch of containers can be launched in a matter of seconds to perform various tasks. The pods of existing Kubernetes clusters can be scheduled to CIS via the Kubernetes API to cope with business surges. CIS is billed based on actual resources used, helping to reduce your computing costs. It can greatly reduce the barriers to deploying containers and reduce the costs associated with performing batch tasks or addressing business spikes.

Benefits

Flexibility and Ease of Use

CIS supports flexible configuration of resources based on container needs and is able to quickly batch launch containers. All you need to do is simply complete the Dockerization of the program which can then be executed by deploying CIS, freeing you from concerning over the computing resources at the IaaS layer in the cloud. By default, when program operation ends, the container will also end. Restart policies can be configured to control the lifecycle of the container.

Security and Reliability

The underlying layer of CIS uses Clear Container technology to achieve virtualized isolation for CVM-grade security. CIS runs on your VPC, supports configuration of security groups and access control lists and ensures network security of the container through custom policies.

Service Integration

CIS can be deeply integrated with most Tencent Cloud services such as CVM, COS, CDB and VPC, enriching the solutions offered by cloud computing services to meet your different business needs.

Low Cost

Containers are charged based on the CPU, memory and network resources consumed from run start to run end with no additional service fees. Additionally, the container lifecycle is almost identical to the lifecycle of the business program running inside it, which means that it is charged based on the amount of resources used for running the business with virtually no waste.

Compatibility with Kubernetes

CIS can be managed by an existing Kubernetes cluster through the upper-layer scheduler. CIS can be scheduled as pods to Tencent Kubernetes Engine (TKE, Tencent Cloud's container service) or your self-built Kubernetes cluster through the Kubernetes API. Creating these CIS-based pods does not require addition of new resources or consumption of existing server resources, and the pods can be managed by the cluster's master.

Features

Rapid Container Deployment

CIS provides a minimalist container experience. Unlike TKE, there is no need to pre-create clusters, purchase CVMs for production containers or have knowledge of container management tools such as Kubernetes. Compared to CVM, CIS does not require server OPS and management, Docker execution environment deployment or operating system compatibility guarantee. Containers can be deployed in Tencent Cloud after simply determining the network environment for container running, container image and version and the CPU and memory size required for normal operation.

Scenarios

Program Hosting

For stateless programs, CIS is an ideal execution environment. As long as the user has an image of the program's container, the program can be run at a very low cost through simple configurations. For example, any simple container-enabled application such as performing a program verification, crawling a website or deploying a web service can be deployed using CIS. In addition, CIS supports carrying stateful programs by specifying a fixed public IP and using persistent cache.

Computing Jobs

CIS is the simplest way to run containers in Tencent Cloud. In addition to its competitive pricing, it can batch launch instances in seconds, release them at the end of the logic and interconnect with various private and public networks and resources, making it ideal for computing jobs. No matter whether it's small-scale algorithm validation or large-scale batch computing, serial task using CPU or parallel task using GPU, CIS can be the perfect execution environment.