It will no longer be maintained. Please migrate to apache/cloudstack-kubernetes-provider |
A Cloud Controller Manager to facilitate Kubernetes deployments on Cloudstack.
Based on the old Cloudstack provider in Kubernetes that will be removed soon.
Prebuilt containers are posted on Docker Hub.
The cloud controller is intended to be deployed as a daemon set, with on instance running on each node.
To configure API access to your CloudStack management server, you need to create a secret containing a cloudstack.ini
that is suitable for your environment.
cloudstack.ini
should look like this:
[Global]
api-url = <CloudStack API URL>
api-key = <CloudStack API Key>
secret-key = <CloudStack API Secret>
project-id = <CloudStack Project UUID (optional)>
zone = <CloudStack Zone Name (optional)>
ssl-no-verify = <Disable SSL certificate validation: true or false (optional)>
The access token needs to be able to fetch VM information and deploy load balancers in the project or domain where the nodes reside.
To create the secret, use the following command:
kubectl create secret generic cloudstack-secret --from-file=cloudstack.ini
You can then use the provided example deployment.yaml to deploy the controller:
kubectl apply -f deployment.yaml
This CCM supports TCP, UDP and TCP-Proxy LoadBalancer deployments.
For UDP and Proxy Protocol support, CloudStack 4.6 or later is required.
Since kube-proxy does not support the Proxy Protocol or UDP, you should connect this directly to pods, for example by deploying a DaemonSet and setting hostNetwork: true
.
The service running in the pod must support the protocol.
See service.yaml for an example Service deployment and part of a suitable configuration for an ingress controller.
It is recommended to launch kubelet
with the following parameter:
--register-with-taints=node.cloudprovider.kubernetes.io/uninitialized=true:NoSchedule
This will treat the node as 'uninitialized' and cause the CCM to apply metadata labels from CloudStack automatically.
Supported labels are:
- kubernetes.io/hostname (= the instance name)
- beta.kubernetes.io/instance-type (= the compute offering)
- failure-domain.beta.kubernetes.io/zone (= the zone)
- failure-domain.beta.kubernetes.io/region (also = the zone)
It is also possible to trigger this process manually by issuing the following command:
kubectl taint nodes <my-node-without-labels> node.cloudprovider.kubernetes.io/uninitialized=true:NoSchedule
There are several notable differences to the old Kubernetes CloudStack cloud provider that need to be taken into account when migrating from the old cloud provider to the standalone controller.
Load balancer rule names now include the protocol in addition to the LB name and service port. This was added to distinguish tcp, udp and tcp-proxy services operating on the same port. Without this change, it would not be possible to map a service that runs on both TCP and UDP port 8000, for example.
If you don't do this, you will end up with duplicate rules for the same service, which won't work.
Since the controller is now intended to be run inside a pod and not on the node, it will not be able to fetch metadata from the Virtual Router's DHCP server.
Instead, it first obtains the name of the node from Kubernetes, then fetches information from the CloudStack API.
Go 1.12 is required to build cloudstack-ccm.
It is recommended to build with Go module support, but dependencies are also vendored.
To build the controller with correct versioning, some build flags need to be passed. A Makefile is provided that sets these build flags to values obtained from git.
go get github.com/apache/cloudstack-kubernetes-provider
cd ${GOPATH}/src/github.com/apache/cloudstack-kubernetes-provider
make
To build the cloudstack-cloud-controller-manager container, please use the provided Dockerfile. The Makefile will also with that and properly tag the resulting container.
make docker
You need a local instance of the CloudStack Management Server or a 'real' one to connect to. The CCM supports the same cloudstack.ini configuration file format used by the cs tool, so you can simply point it to that.
./cloudstack-ccm --cloud-provider external-cloudstack --cloud-config ~/.cloudstack.ini --master k8s-apiserver
Replace k8s-apiserver with the host name of your Kubernetes development clusters's API server.
If you don't have a 'real' CloudStack installation, you can also launch a local simulator instance instead. This is very useful for dry-run testing.
© 2018 SWISS TXT AG and the Kubernetes authors.
See LICENSE-2.0 for permitted usage.