Elastic Container Instance:Create a pod by specifying a compute category

Economy

Supported in all regions where ECI is available

Supported only in some regions

General

Supported in all regions where ECI is available

Supported only in some regions

Step 1: Specify vCPU and memory

Before setting a compute category, specify the pod's vCPU and memory using one of these methods:

If none of limits, requests, or k8s.aliyun.com/eci-use-specs are set, or if all three are set simultaneously, see Configuration description for how the system resolves the spec.

Step 2: Set the compute category

Add the k8s.aliyun.com/eci-compute-category annotation to the pod's metadata section.

• Valid values: economy, general

• To specify a fallback order, use a comma-separated list: general,economy

Multiple specs and categories

When you specify multiple values for both, the system iterates through all combinations by spec first, then category. Invalid combinations (for example, 16 vCPU + economy) are automatically skipped.

Example: k8s.aliyun.com/eci-use-specs: "4-8Gi,8-16Gi,16-32Gi" and k8s.aliyun.com/eci-compute-category: "general,economy"

The system tries in this order:

1. 4 vCPU / 8 GiB, general

2. 4 vCPU / 8 GiB, economy

3. 8 vCPU / 16 GiB, general

4. 8 vCPU / 16 GiB, economy

5. 16 vCPU / 32 GiB, general

6. 16 vCPU / 32 GiB, economy *(invalid — auto-skipped)*

Example 1: Economy category with per-container limits

Use limits on each container and set the compute category to economy. The pod's total vCPU and memory are summed across containers.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: test
  labels:
    app: test
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      name: test
      labels:
        app: nginx
        alibabacloud.com/eci: "true"
      annotations:
        k8s.aliyun.com/eci-compute-category: "economy"
    spec:
      containers:
      - name: nginx
        image: registry.cn-shanghai.aliyuncs.com/eci_open/nginx:1.14.2
        ports:
        - containerPort: 80
        resources:
          limits:
            cpu: "500m"       # 0.5 vCPU
            memory: "1024Mi"  # 1 GiB
      - name: busybox
        image: registry.cn-shanghai.aliyuncs.com/eci_open/busybox:1.30
        command: ["sleep"]
        args: ["999999"]
        resources:
          limits:
            cpu: "1000m"      # 1 vCPU
            memory: "2048Mi"  # 2 GiB

Example 2: General with economy fallback

Set general,economy to prefer high performance but fall back to economy if general resources are unavailable. This improves availability without requiring a fixed compute category.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: test
  labels:
    app: test
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      name: test
      labels:
        app: nginx
        alibabacloud.com/eci: "true"
      annotations:
        k8s.aliyun.com/eci-compute-category: "general,economy"  # Prefer general; fall back to economy
    spec:
      containers:
      - name: nginx
        image: registry.cn-shanghai.aliyuncs.com/eci_open/nginx:1.14.2
        ports:
        - containerPort: 80
        resources:
          limits:
            cpu: "2000m"      # 2 vCPUs
            memory: "4096Mi"  # 4 GiB
      - name: busybox
        image: registry.cn-shanghai.aliyuncs.com/eci_open/busybox:1.30
        command: ["sleep"]
        args: ["999999"]
        resources:
          limits:
            cpu: "2000m"      # 2 vCPUs
            memory: "4096Mi"  # 4 GiB

Example 3: Multiple specs with economy category

Use k8s.aliyun.com/eci-use-specs to specify multiple pod-level specs with a fallback order, combined with a single compute category.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: test
  labels:
    app: test
spec:
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      name: test
      labels:
        app: nginx
        alibabacloud.com/eci: "true"
      annotations:
        k8s.aliyun.com/eci-use-specs: "2-4Gi,4-8Gi"    # Try 2 vCPU/4 GiB first, then 4 vCPU/8 GiB
        k8s.aliyun.com/eci-compute-category: "economy"
    spec:
      containers:
      - name: nginx
        image: registry.cn-shanghai.aliyuncs.com/eci_open/nginx:1.14.2
        ports:
        - containerPort: 80