Assign IP addresses based on topology

    Value

    If you have workloads in different regions, zones, or rack, you may want them to get IP addresses from the same IP pool. This strategy is useful for reducing the number of routes that are required in the network, or to meet requirements imposed by an external firewall device or policy. Calico makes it easy to do this using an IP pool resource with node labels and node selectors.

    This how-to guide uses the following Calico features:

    • IPPool resource

    Concepts

    Topology-based IP address assignment requires addresses to be per-host (node). As such, Kubernetes annotations cannot be used because annotations are only per-namespace and per-pod. And although you can configure IP addresses for nodes in the CNI configuration, you are making changes within the host’s file system. The best option is to use node-selection IP address assignment using IP pools.

    Node selection-based IP address assignment is exactly what it sounds like: node labels are set, and Calico uses node selectors to decide which IP pools to use when assigning IP addresses to the node.

    Nodes only assign workload addresses from IP pools which select them. To avoid having a workload not get an IP and fail to start, it is important to ensure that all nodes are selected by at least one IP pool.

    In the following example, we create an IP pool that only allocates IP addresses for nodes with the label, zone=west.

    Then, we label a node with zone=west. For example:

    Tutorial

    1.        -------------------
    2.        |    router       |
    3.        -------------------
    4.        |                 |
    5. | rack-0      |   | rack-1      |
    6. ---------------   ---------------
    7. | kube-node-0 |   | kube-node-2 |
    8. - - - - - - - -   - - - - - - - -
    9. | kube-node-1 |   | kube-node-3 |
    10. - - - - - - - -   - - - - - - - -

    Using the pod IP range 192.168.0.0/16, we target the following setup: reserve the 192.168.0.0/24 and 192.168.1.0/24 pools for rack-0, . Let’s get started.

    By installing Calico without setting the default IP pool to match, running calicoctl get ippool -o wide shows that Calico created its default IP pool of 192.168.0.0/16:

    1. Delete the default IP pool.

      Since the default-ipv4-ippool IP pool resource already exists and accounts for the entire /16 block, we will have to delete this first:

      1. calicoctl delete ippools default-ipv4-ippool

    2. Label the nodes.

      To assign IP pools to specific nodes, these nodes must be labelled using kubectl label.

      1. kubectl label nodes kube-node-0 rack=0
      2. kubectl label nodes kube-node-1 rack=0
      3. kubectl label nodes kube-node-2 rack=1
      4. kubectl label nodes kube-node-3 rack=1

    3. Create an IP pool for each rack.

      1. calicoctl create -f -<<EOF
      2. apiVersion: projectcalico.org/v3
      3. kind: IPPool
      4.   name: rack-1-ippool
      5. spec:
      7.   ipipMode: Always
      8.   natOutgoing: true
      9.   nodeSelector: rack == "1"
      10. EOF
      1. NAME                  CIDR             NAT    IPIPMODE   DISABLED   SELECTOR
      2. rack-0-ippool         192.168.0.0/24   true   Always     false      rack == "0"
      3. rack-1-ippool         192.168.1.0/24   true   Always     false      rack == "1"

    4. Verify that the IP pool node selectors are being respected.

      We will create an nginx deployment with five replicas to get a workload running on each node.

      Check that the new workloads now have an address in the proper IP pool allocated for the rack that the node is on with kubectl get pods -owide.

      1. NAME                   READY   STATUS    RESTARTS   AGE    IP             NODE          NOMINATED NODE   READINESS GATES
      2. nginx-5c7588df-prx4z   1/1     Running   0          6m3s   192.168.0.64   kube-node-0   <none>           <none>
      3. nginx-5c7588df-s7qw6   1/1     Running   0          6m7s   192.168.0.129  kube-node-1   <none>           <none>
      4. nginx-5c7588df-w7r7g   1/1     Running   0          6m3s   192.168.1.65   kube-node-2   <none>           <none>
      5. nginx-5c7588df-62lnf   1/1     Running   0          6m3s   192.168.1.1    kube-node-3   <none>           <none>
      6. nginx-5c7588df-pnsvv   1/1     Running   0          6m3s   192.168.1.64   kube-node-2   <none>           <none>

      The grouping of IP addresses assigned to the workloads differ based on what node that they were scheduled to. Additionally, the assigned address for each workload falls within the respective IP pool that selects the rack that they run on.

    note

    Calico IPAM will not reassign IP addresses to workloads that are already running. To update running workloads with IP addresses from a newly configured IP pool, they must be recreated. We recommend doing this before going into production or during a maintenance window.

    Calico IPAM