Alibaba Cloud Container Compute Service (ACS) is a cloud computing service that provides container compute resources complying with the container specifications of Kubernetes. ACS provides serverless container computer resources and eliminates the need to worry about node and cluster O&M. ACS applies to a variety of use scenarios to support the workloads of containerized applications and cloud services.
Product introduction
What is Container Compute Service (ACS)?
Container Compute Service (ACS) is an upgrade of ACK Serverless clusters (FKA ASK). It is more cost-effective, easy-to-use, and elastic. ACS is intended for a variety of business scenarios. It defines cost-effective serverless compute classes and compute QoS classes, allowing you to request resources on demand and pay for them on a per-second basis. It saves you the need to worry about cluster and node O&M.
Comparison item | ACS cluster | ACK Pro cluster |
Comparison item | ACS cluster | ACK Pro cluster |
Features |
|
|
Billing | No cluster management fee is charged. Pay-as-you-go fees are charged on a per-second basis based on the instance type of the pods that are created and the vCPU, memory, or GPU resource allocated to the pods. For more information, see Billing. | A cluster management fee is charged. Nodes are billed based on specifications and uptime. For more information, see Billing overview. |
ACS compute classes
ACS defines compute classes for CPU compute power and GPU compute power to help you reduce the complexity of resource allocation and cluster sizing. For more information, see Overview of ACS pods.
Compute class: general-purpose | Compute class: performance-enhanced | Compute class: GPU-accelerated | GPU-HPN |
Compute class: general-purpose | Compute class: performance-enhanced | Compute class: GPU-accelerated | GPU-HPN | |
QoS class: default |
|
|
|
|
QoS class: best-effort |
|
|
|
|
GPU compute power is in invitational preview. To experience GPU compute power, contact the sales manager or PDSA.
Benefits
Cost-effectiveness and ease of use
ACS provides general-purpose and performance-enhanced pods for online businesses. ACS provides cost-effective BestEffort pods for offline businesses. You can use YAML templates or the ACS console to quickly create and deploy pods.
Fine-grained and elastic resource requests
For CPU compute power, you can request a minimum of 0.25 vCPUs and 0.5 GiB of memory, and increase the resource request with a step size of 0.5 vCPUs and 1 GiB of memory. For GPU compute power, you can request a minimum of one GPU. This enables you to apply for resources on demand to reduce the expenses.
On-demand scaling and pay-as-you-go billing
You can scale large-scale workloads within seconds and apply for elastic resources on demand. Pay-as-you-go fees are charged on a per-second basis. ACS allows you to purchase savings plans based on daily committed consumption to handle workload fluctuations with reduced costs.
Simplicity and rich scenarios
ACS hosts the key system components of Kubernetes clusters and supports auto updates to cluster patch versions. This greatly reduces the complexity of cluster O&M. ACS applies to various business scenarios and supports the workloads of containerized applications and cloud services.
Use scenarios
Common online businesses
For common online applications such as microservices applications and web applications, you can use cost-effective and stable general-purpose pods. You can launch pods and scale large-scale workloads within seconds to handle unexpected traffic spikes. This prevents business loss and avoids resource waste caused by resource scale-out.
Big data computing businesses
For big data computing businesses that are not sensitive to the latency but require high data throughput, such as Spark, Presto, and AI training jobs, you can use cost-effective BestEffort pods. You can launch pods and scale large-scale workloads within seconds to improve the efficiency of parallel computing.
AI training and inference businesses
For AI inference businesses that are sensitive to latency and deployed in real time, such as AIGC model training and inference, autonomous driving model training and inference, and on-cloud real-time inference, you can use a combination of general-purpose GPU-accelerated pods and GPU capacity reservations/GPU-HPN capacity reservations to guarantee resource supply and reduce costs.
High-performance businesses
ACS is also suitable for high-performance businesses, such as cloud gaming. You can launch pods and scale large-scale workloads within seconds to handle unexpected traffic spikes and ensure the optimal user experience. This improves the processing speed and further reduces the response latency and stutters.
Key features
Resource management
Feature | Description |
Feature | Description |
Pod types | ACS provides general-purpose, performance-enhanced, GPU-accelerated, and GPU-HPN pods to meet requirements in different business scenarios. For more information, see Overview of ACS pods. |
On-demand scaling | By default, ACS scales resources on demand and requests resources based on the type of pod. Pay-as-you-go fees are charged on a per-second basis. You can view the metering data on the billing details page. |
Capacity reservations | ACS provides pod capacity reservations and node capacity reservations for pods that run GPU inference businesses. |
Cluster management
Feature | Description |
Feature | Description |
Cluster creation | ACS clusters use serverless resources, which saves you the need to maintain clusters and nodes. You can use YAML templates or the ACS console to quickly create and deploy workloads. This greatly narrows the technical gap for using Kubernetes and Alibaba Cloud services. |
Cluster connection | You can obtain the kubeconfig file of an ACS cluster and then use kubectl to connect to the ACS cluster. This allows you to easily manage and use ACS clusters and ACS resources. |
Authorization management | ACS allows you to manage RAM permissions for resources and RBAC permissions for Kubernetes clusters. For more information, see Authorization overview. |
Scheduling domain management | ACS is compatible with Kubernetes scheduling, supports colocation of multiple types of workloads, and provides fine-grained scheduling for elastic resources and heterogeneous resources. |
After you log on to the ACK console, you can view ACS clusters, manage RBAC permissions for ACS clusters, and manage the kubeconfig files of ACS clusters.
Application management
Feature | Description |
Feature | Description |
Application creation | ACS supports a variety of workloads, including Deployments, StatefulSets, Jobs, and CronJobs. You can create workloads from a client, in the console, or by using a template. You can also configure environment variables, health check, data disks, and logging when you create workloads. |
Application scaling | |
Storage management | Volumes used in ACS clusters are managed by the CSI plug-in and support Elastic Block Storage (EBS) and Apsara File Storage NAS (NAS). |
Network management | ACS provides stable and high-performance container networks by integrating the Kubernetes network model, Virtual Private Cloud (VPC), and Server Load Balancer (SLB). |
Security and O&M
Category | Feature | Description |
Category | Feature | Description |
Observability | Monitoring | Managed Service for Prometheus is integrated with ACS and enabled by default. With the built-in dashboards and performance metrics, you can monitor the status of Kubernetes clusters, pods, and applications from multiple dimensions. |
Logs | ACS is integrated with Simple Log Service, allowing you to collect and view application logs, pod logs, and cluster logs. | |
Alerting | You can configure alerts to manage exceptions in clusters based on various metrics for different scenarios. | |
Cluster inspection | Cluster inspection | ACS provides the cluster inspection feature to automatically scan for potential risks in clusters. |
Diagnostics | ACS provides the diagnostics feature to diagnose pods, Services, and Ingresses. | |
Security center | Audit | ACS generates API server audit logs to help you record or trace the daily operations of users. |
Limits on ACS clusters
Before you use Container Compute Service (ACS), take note of the following limits on ACS clusters.
ACS clusters do not support DaemonSets. You can replace DaemonSets with sidecar containers.
You cannot specify
HostPathorHostNetworkin podmanifests.ACS clusters do not support privileged containers. You can use a security context to add capabilities to a pod.
ACS clusters do not support NodePort Services or session affinity.
ACS clusters do not support the China East Finance, China South Finance, or Alibaba Gov Cloud regions.
References
For more information about how to activate and use ACS, see Quick start for first-time ACS users.
To view the release notes of ACS, see Release notes.
For more information about solutions used in different scenarios based on ACS, see Best practices.
