Calico charm
Calico is a new approach to virtual networking and network security for containers, VMs, and bare metal services, that provides a rich set of security enforcement capabilities running on top of a highly scalable and efficient virtual network fabric.
This charm will deploy calico as a background service, and configure CNI for use with calico, on any principal charm that implements the kubernetes-cni interface.
Usage
The calico charm is a subordinate. This charm will require a principal charm
that implements the kubernetes-cni interface in order to properly deploy.
juju deploy cs:~containers/calico
juju deploy cs:~containers/etcd
juju deploy cs:~containers/kubernetes-master
juju deploy cs:~containers/kubernetes-worker
juju add-relation calico etcd
juju add-relation calico kubernetes-master
juju add-relation calico kubernetes-worker
| name | type | Default | Description |
|---|---|---|---|
| bgp-service-cluster-ips | string | See notes | |
| bgp-service-external-ips | string | See notes | |
| calico-node-image | string | See notes | The image id to use for calico/node. |
| calico-policy-image | string | See notes | The image id to use for calico/kube-controllers. |
| cidr | string | 192.168.0.0/16 | Network CIDR assigned to Calico. This is applied to the default Calico pool, and is also communicated to the Kubernetes charms for use in kube-proxy configuration. |
| disable-vxlan-tx-checksumming | boolean | True | See notes |
| global-as-number | int | 64512 | Global AS number. |
| global-bgp-peers | string | [] | List of global BGP peers. Each BGP peer is specified with an address and an as-number. Example value: "[{address: 10.0.0.1, as-number: 65000}, {address: 10.0.0.2, as-number: 65001}]" |
| ignore-loose-rpf | boolean | False | Enable or disable IgnoreLooseRPF for Calico Felix. This is only used when rp_filter is set to a value of 2. |
| ipip | string | Never | IPIP encapsulation mode. Must be one of "Always", "CrossSubnet", or "Never". This is incompatible with VXLAN encapsulation. If VXLAN encapsulation is enabled, then this must be set to "Never". |
| manage-pools | boolean | True | If true, a default pool is created using the cidr and ipip charm configuration values. Warning: When manage-pools is enabled, the charm will delete any pools that are unrecognized. |
| nat-outgoing | boolean | True | NAT outgoing traffic |
| node-to-node-mesh | boolean | True | When enabled, each Calico node will peer with every other Calico node in the cluster. |
| route-reflector-cluster-ids | string | {} | Mapping of unit IDs to route reflector cluster IDs. Assigning a route reflector cluster ID allows the node to function as a route reflector. Example value: "{0: 224.0.0.1, 2: 224.0.0.1}" |
| subnet-as-numbers | string | {} | See notes |
| subnet-bgp-peers | string | {} | See notes |
| unit-as-numbers | string | {} | See notes |
| unit-bgp-peers | string | {} | See notes |
| veth-mtu | int | None | See notes |
| vxlan | string | Never | VXLAN encapsulation mode. Must be one of "Always", "CrossSubnet", or "Never". This is incompatible with IPIP encapsulation. If IPIP encapsulation is enabled, then this must be set to "Never". |
bgp-service-cluster-ips
Space-separated list of service cluster CIDRs to advertise over BGP. These will be passed to the .spec.serviceClusterIPs field of the default BGPConfiguration in Calico.
Example value: ”10.0.0.0/24 10.0.1.0/24”
bgp-service-external-ips
Space-separated list of service external CIDRs to advertise over BGP. These will be passed to the .spec.serviceExternalIPs field of the default BGPConfiguration in Calico.
Example value: ”10.0.0.0/24 10.0.1.0/24”
calico-node-image
rocks.canonical.com:443/cdk/calico/node:v3.10.1
calico-policy-image
rocks.canonical.com:443/cdk/calico/kube-controllers:v3.10.1
disable-vxlan-tx-checksumming
When set to true, if VXLAN encapsulation is in use, then the charm will disable TX checksumming on the vxlan.calico network interface. This works around an upstream issue in Calico: https://github.com/projectcalico/calico/issues/3145
subnet-as-numbers
Mapping of subnets to AS numbers, specified as YAML. Each Calico node will be assigned an AS number based on the entries in this mapping.
Example value: "{10.0.0.0/24: 64512, 10.0.1.0/24: 64513}"
If a node's IP matches any of the specified subnets, then the corresponding AS number is used instead of the global one.
If a node's IP matches no subnets, then the global AS number will be used instead.
If a node's IP matches multiple subnets, then the most specific subnet will be used, e.g. a /24 subnet will take precedence over a /16.
subnet-bgp-peers
Mapping of subnets to lists of BGP peers. Each BGP peer is specified with an address and an as-number.
Example value: "{10.0.0.0/24: [{address: 10.0.0.1, as-number: 65000}, {address: 10.0.0.2, as-number: 65001}], 10.0.1.0/24: [{address: 10.0.1.1, as-number: 65002}]}"
If a node's IP matches multiple subnets, then peerings will be added for each matched subnet.
unit-as-numbers
Mapping of unit IDs to AS numbers, specified as YAML. Each Calico node will be assigned an AS number based on the entries in this mapping.
Example value: "{0: 64512, 1: 64513}"
This takes precedence over global-as-number and subnet-as-numbers.
unit-bgp-peers
Mapping of unit IDs to lists of BGP peers. Each BGP peer is specified with an address and an as-number.
Example value: "{0: [{address: 10.0.0.1, as-number: 65000}, {address: 10.0.0.2, as-number: 65001}], 1: [{address: 10.0.1.1, as-number: 65002}]}"
veth-mtu
Set veth MTU size. This should be set to the MTU size of the base network.
If VXLAN is enabled, then the charm will automatically subtract 50 from the specified MTU size.
If IPIP is enabled, then the charm will automatically subtract 20 from the specified MTU size.