Configuring an EIP for a Pod¶
Application Scenarios¶
In Cloud Native Network 2.0, pods use VPC ENIs or sub-ENIs for networking. You can directly bind EIPs to pods.
To associate an EIP with a pod, simply set the value of the yangtse.io/pod-with-eip annotation to true when creating the pod. Then, the EIP will be automatically allocated and bound to the pod.
Prerequisites¶
A CCE Turbo cluster is available and the cluster version meets the requirements listed in the following table.
Scenario | Cluster Version |
|---|---|
Allocating an EIP with a pod |
|
Using an existing EIP for a pod | v1.23: v1.23.16-r0 or later v1.25: v1.25.11-r0 or later v1.27: v1.27.8-r0 or later v1.28: v1.28.6-r0 or later v1.29: v1.29.2-r0 or later Versions later than v1.29 |
Notes and Constraints¶
To access a pod bound with an EIP from the Internet, add security group rules to allow the target request traffic.
Only one EIP can be bound to a pod.
Configure the EIP-related annotation when creating a pod. After the pod is created, the annotations related to the EIP cannot be modified.
Do not perform operations on the EIP associated with a pod through the EIP console or API. Otherwise, the EIP may malfunction. The operations include changing the EIP name, deleting, and unbinding or binding the EIP.
After an automatically allocated EIP is manually deleted, the network malfunctions. In this case, rebuild the pod.
Allocating an EIP with a Pod¶
If you set the pod-with-eip annotation to true when creating a pod, an EIP will be automatically allocated and bound to the pod.
Note
CCE will automatically add cluster ID, namespace, and pod name labels to an EIP that has been allocated automatically.
If a pod is created with an automatically allocated EIP, deleting the pod will also delete the EIP.
The following uses a Deployment named nginx as an example. For details about annotations, see Table 2.
For an automatically allocated EIP with a dedicated bandwidth when you create a Deployment, you do not need to specify the bandwidth ID. The following shows an example:
apiVersion: apps/v1 kind: Deployment metadata: name: nginx spec: replicas: 3 selector: matchLabels: app: nginx template: metadata: labels: app: nginx annotations: yangtse.io/pod-with-eip: "true" # An EIP will be automatically allocated when the pod is created. yangtse.io/eip-bandwidth-size: "5" # EIP bandwidth yangtse.io/eip-network-type: 5_bgp # EIP type yangtse.io/eip-charge-mode: bandwidth # EIP billing mode yangtse.io/eip-bandwidth-name: <eip_bandwidth_name> # EIP bandwidth name spec: containers: - name: container-0 image: nginx:alpine resources: limits: cpu: 100m memory: 200Mi requests: cpu: 100m memory: 200Mi imagePullSecrets: - name: default-secret
Table 1 Annotations of an EIP with a dedicated bandwidth¶ Annotation
Mandatory
Default Value
Description
Value Range
yangtse.io/pod-with-eip
Yes
false
Whether to allocate an EIP with a pod and bind the EIP to the pod
false or true
yangtse.io/eip-bandwidth-size
No
5
Bandwidth, in Mbit/s
The value range varies depending on the region and bandwidth billing mode. For details, see the EIP console.
yangtse.io/eip-network-type
No
5_bgp
EIP type
The type varies depending on the region. For details, see on the EIP console.
yangtse.io/eip-charge-mode
No
None
Billed by traffic or bandwidth
You are advised to configure this parameter. If this parameter is left blank, no billing mode is specified. In this case, the default value of the EIP API in the region is used.
bandwidth: billed by bandwidth
traffic: billed by traffic
yangtse.io/eip-bandwidth-name
No
Pod name
Bandwidth name
Enter 1 to 64 characters. Only letters, digits, hyphens (-), underscores (_), and periods (.) are allowed.
Minimum length: 1 character
Maximum length: 64 characters
For an automatically allocated EIP with a shared bandwidth when you create a Deployment, you are required to specify the bandwidth ID. The following shows an example:
apiVersion: apps/v1 kind: Deployment metadata: name: nginx spec: replicas: 3 selector: matchLabels: app: nginx template: metadata: labels: app: nginx annotations: yangtse.io/pod-with-eip: "true" # An EIP will be automatically allocated when the pod is created. yangtse.io/eip-network-type: 5_bgp # EIP type yangtse.io/eip-bandwidth-id: <eip_bandwidth_id> # Shared bandwidth ID of the EIP spec: containers: - name: container-0 image: nginx:alpine resources: limits: cpu: 100m memory: 200Mi requests: cpu: 100m memory: 200Mi imagePullSecrets: - name: default-secret
Table 2 Annotations of an EIP with a shared bandwidth¶ Annotation
Mandatory
Default Value
Description
Value Range
yangtse.io/pod-with-eip
Yes
false
Whether to allocate an EIP with a pod and bind the EIP to the pod
false or true
yangtse.io/eip-network-type
No
5_bgp
EIP type
5_bgp
The specific type varies with regions. For details, see the EIP console.
yangtse.io/eip-bandwidth-id
Mandatory when a shared bandwidth is used
None
ID of an existing bandwidth
If this parameter is not specified, the EIP with a dedicated bandwidth is used by default. For details about how to configure parameters for an EIP with a dedicated bandwidth, see Table 1.
Only the yangtse.io/eip-network-type field can be specified concurrently, and this field is optional.
None
Using an Existing EIP for a Pod¶
If you configure the yangtse.io/eip-id annotation when creating a pod, the EIP will be automatically bound to the pod.
Note
After an existing EIP is bound to a pod, CCE will automatically add the cluster ID, namespace, and pod name labels to the EIP.
When a pod using an existing EIP is deleted, the EIP will be retained. CCE will automatically delete the labels for the cluster ID, namespace, and pod name that were added when the EIP was bound.
The following example shows how to create a Deployment named nginx with a single pod with an EIP automatically bound.
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx
spec:
replicas: 1
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
annotations:
yangtse.io/eip-id: <eip_id> # ID of an existing EIP
spec:
containers:
- name: container-0
image: nginx:alpine
resources:
limits:
cpu: 100m
memory: 200Mi
requests:
cpu: 100m
memory: 200Mi
imagePullSecrets:
- name: default-secret
Annotation | Mandatory | Description |
|---|---|---|
yangtse.io/eip-id | Yes | EIP ID How to obtain: Log in to the EIP console, click the name of the target EIP in the EIP list, and copy the ID field. |
Checking Whether the EIP Bound to the Pod Is Available¶
After an EIP is allocated to a pod, the container networking controller binds the EIP to the pod and writes the allocation result back to the pod's yangtse.io/allocated-ipv4-eip annotation. The startup time of the pod's service containers may be earlier than the time when the EIP allocation result is written back.
You can configure an init container for the pod, associate the yangtse.io/allocated-ipv4-eip annotation with the init container through a downwardAPI volume, and check whether the EIP has been allocated in the init container. You can configure the init container as follows:
apiVersion: v1
kind: Pod
metadata:
name: example
annotations:
yangtse.io/pod-with-eip: "true"
yangtse.io/eip-bandwidth-size: "5"
yangtse.io/eip-network-type: 5_bgp
yangtse.io/eip-charge-mode: bandwidth
yangtse.io/eip-bandwidth-name: "xxx"
spec:
initContainers:
- name: init
image: busybox:latest
command: ['timeout', '60', 'sh', '-c', "until grep -E '[0-9]+' /etc/eipinfo/allocated-ipv4-eip; do echo waiting for allocated-ipv4-eip; sleep 2; done"]
volumeMounts:
- name: eipinfo
mountPath: /etc/eipinfo
volumes:
- name: eipinfo
downwardAPI:
items:
- path: "allocated-ipv4-eip"
fieldRef:
fieldPath: metadata.annotations['yangtse.io/allocated-ipv4-eip']
...