Container Service for Kubernetes:Configure bandwidth limits for pods

Prerequisites

• An existing ACK Managed Cluster or ACK Dedicated Cluster that uses the Flannel network plugin. For instructions, see Create an ACK managed cluster or Create an ACK dedicated cluster (discontinued). To learn how to install the Flannel network plugin when you create a cluster, see Work with Flannel.

• Your cluster must have Flannel network plugin v0.15.1.4-e02c8f12-aliyun or later. Check the plugin version on the Components page. For information about how to access the Components page, see Manage components.

  Note

  If you cannot upgrade the Flannel component to the required version, first upgrade the Kubernetes version of your cluster. For instructions, see Manually upgrade a cluster.

Configure bandwidth limits

To enable bandwidth limiting, you must modify the Flannel configuration file. If you have not yet enabled it, see Enable the bandwidth limit feature for an ACK cluster.

The Flannel network plugin lets you control the network bandwidth of a pod. Use the following pod annotations to specify the maximum inbound and outbound bandwidth:

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: demo
  name: demo
  namespace: default
spec:
  selector:
    matchLabels:
      app: demo
  template:
    metadata:
      annotations:
        # Limit the maximum inbound bandwidth of the pod to 10 Mbps.
        kubernetes.io/ingress-bandwidth: 10M
        # Limit the maximum outbound bandwidth of the pod to 30 Mbps.
        kubernetes.io/egress-bandwidth: 30M
      labels:
        app: demo
    spec:
      containers:
      # The rest of the configuration is omitted.

Enable the bandwidth limit feature for an ACK cluster

1. Run the following command to open and edit the ConfigMap for the Flannel component.

   kubectl edit cm -n kube-system kube-flannel-cfg

2. Under data.cni-conf.json, add the bandwidth configuration at the end of the plugins array. Then, save and exit the file.

   Note

   Ensure the content is valid JSON. Use only single-byte characters and verify there are no missing or extra commas.

   apiVersion: v1
   data:
     cni-conf.json: |
       {
         "name": "cb0",
         "cniVersion":"0.3.1",
         "plugins": [
           {
             "type": "flannel",
             "delegate": {
               "isDefaultGateway": true,
               "hairpinMode": true
              },
              "dataDir": "/var/run/cni/flannel",
              "ipam": {
                "type": "host-local",
                "dataDir": "/var/run/cni/networks"
              }
           },
           {
             "type": "portmap",
             "capabilities": {
               "portMappings": true
             },
             "externalSetMarkChain": "KUBE-MARK-MASQ"
           },
           {
             "type": "bandwidth",
             "capabilities": {
               "bandwidth": true
             }
           }
         ]
       }

3. Run the following command to delete all Flannel pods. The system automatically recreates them, and the new configuration takes effect.

   This process does not affect running services.

   kubectl -n kube-system delete pod -l app=flannel

4. Run the following command to ensure that all Flannel pods are in the Running state.

   kubectl -n kube-system get pod -o wide -l app=flannel

   The following output indicates that the Flannel pods are running correctly.

   NAME                    READY   STATUS    RESTARTS   AGE   IP              NODE                        NOMINATED NODE   READINESS GATES
   kube-flannel-ds-h45zj   1/1     Running   0          67s   192.XX.XX.118   cn-hangzhou.192.XX.XX.118   <none>           <none>
   kube-flannel-ds-mvfcw   1/1     Running   0          67s   192.XX.XX.119   cn-hangzhou.192.XX.XX.119   <none>           <none>

5. Run the following command to check the logs for each Flannel pod and ensure that there are no errors.

   kubectl -n kube-system logs <pod_name>

   The following output indicates that the Flannel pods are running correctly.

   Defaulted container "kube-flannel" out of: kube-flannel, install-cni-plugin (init), install-cni (init)
   I0925 07:20:53.794715       1 main.go:219] CLI flags config: {etcdEndpoints:http://127.XX.XX.1:4001,http://127.XX.XX.1:2379 etcdPrefix:/coreos.com/network etcdKeyfile: etcdCertfile: etcdCAFile: etcdUsername: etcdPassword: help:false version:false autoDetectIPv4:false autoDetectIPv6:false kubeSubnetMgr:true kubeApiUrl: kubeAnnotationPrefix:flannel.alpha.coreos.com kubeConfigFile: iface:[] ifaceRegex:[] ipMasq:true subnetFile:/run/flannel/subnet.env subnetDir: publicIP: publicIPv6: subnetLeaseRenewMargin:60 healthzIP:0.0.0.0 healthzPort:0 charonExecutablePath: charonViciUri: iptablesResyncSeconds:5 iptablesForwardRules:true ipforwardResyncSeconds:600 netConfPath:/etc/kube-flannel/net-conf.json setNodeNetworkUnavailable:false}
   W0925 07:20:53.794782       1 client_config.go:608] Neither --kubeconfig nor --master was specified.  Using the inClusterConfig.  This might not work.
   I0925 07:20:53.897149       1 kube.go:121] Node controller skips sync
   I0925 07:20:53.897176       1 main.go:239] Created subnet manager: Kubernetes Subnet Manager - cn-hangzhou.192.XX.XX.118