aws-actuator

README

AWS actuator

The command allows to directly interact with the aws actuator.

To build the aws-actuator binary:

go build -o bin/aws-actuator sigs.k8s.io/cluster-api-provider-aws/cmd/aws-actuator

Prerequisities

All the machine manifests assume existence for various AWS resources such as vpc, subnet, security groups, etc. In case you are starting from scratch, you can run hack/aws-provision.sh to create all the required resources:

$ cd hack
$ ENVIRONMENT_ID=UNIQUE_ID ./aws-provision.sh install

where UNIQUE_ID is unique identification of your environment.

Don't forget to run ./aws-provision.sh destroy once you are done.

Create aws instance based on machine manifest

The examples/userdata.yml secret encodes the following user data:

#!/bin/bash
echo "Ahoj" > /tmp/test

$ ./bin/aws-actuator create -m examples/machine-with-user-data.yaml -c examples/cluster.yaml -u examples/userdata.yml --environment-id UNIQUE_ID
DEBU[0000] Describing AMI ami-a9acbbd6                   example=create-machine machine=test/aws-actuator-testing-machine
Machine creation was successful! InstanceID: i-027681ebf9a842183

Once the aws instance is created you can run $ cat /tmp/test to verify it contains the Ahoj string.

Test if aws instance exists based on machine manifest

$ ./bin/aws-actuator exists -m examples/machine-with-user-data.yaml -c examples/cluster.yaml --environment-id UNIQUE_ID
DEBU[0000] checking if machine exists                    example=create-machine machine=test/aws-actuator-testing-machine
DEBU[0000] instance exists as "i-027681ebf9a842183"      example=create-machine machine=test/aws-actuator-testing-machine
Underlying machine's instance exists.

Delete aws instance based on machine manifest

$ ./bin/aws-actuator delete -m examples/machine-with-user-data.yaml -c examples/cluster.yaml --environment-id UNIQUE_ID
WARN[0000] cleaning up extraneous instance for machine   example=create-machine instanceID=i-027681ebf9a842183 launchTime="2018-08-18 15:50:54 +0000 UTC" machine=test/aws-actuator-testing-machine state=running
INFO[0000] terminating instance                          example=create-machine instanceID=i-027681ebf9a842183 machine=test/aws-actuator-testing-machine
Machine delete operation was successful.

Bootstrapping kubernetes cluster with kubeadm via user data

1. Bootstrap the control plane

./bin/aws-actuator create -m examples/master-machine.yaml -c examples/cluster.yaml -u examples/master-userdata.yaml --environment-id UNIQUE_ID

By default networking is enabled on the master machine. You can interact with the master by copying the kubeconfig and using kubectl. You'll need to find the external IP address of your master and then:

ssh ec2-user@ec2-184-73-119-192.compute-1.amazonaws.com "sudo cat /etc/kubernetes/admin.conf" > kubeconfig
export KUBECONFIG=$PWD/kubeconfig
kubectl config set-cluster kubernetes --server=https://ec2-184-73-119-192.compute-1.amazonaws.com:8443
kubectl get nodes
NAME                           STATUS    ROLES     AGE       VERSION
ip-172-31-34-42.ec2.internal   Ready     master    6m        v1.11.2

1. Once the ip address of the master node is known (e.g. 172.31.34.2), update the kubeadm join line in examples/worker-user-data.sh to:

kubeadm join 172.31.34.2:8443 --token 2iqzqm.85bs0x6miyx1nm7l --discovery-token-unsafe-skip-ca-verification

You can get the internal IP address of the cluster dynamically by running:

echo $(ssh ec2-user@ec2-184-73-119-192.compute-1.amazonaws.com wget -qO - http://169.254.169.254/latest/meta-data/local-ipv4)
172.31.34.42

1. Encode the examples/worker-user-data.sh by running:

$ cat examples/worker-user-data.sh | base64

1. Update the data.userData of examples/worker-user-data.yaml with the generated hash

2. Create the worker node by running:

./bin/aws-actuator create -m examples/worker-machine.yaml -c examples/cluster.yaml -u examples/worker-userdata.yaml --environment-id UNIQUE_ID

After some time the kubernetes cluster with the control plane (master node) and the worker node gets provisioned and the worker joins the cluster.

Bootstrapping cluster API stack

The following command will deploy kubernetes cluster with cluster API stack deployed inside. Worker nodes are deployed with a machineset. It's assumed both AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY are set. It takes some time before the worker node joins the cluster (~3 minutes).

$ ./bin/aws-actuator bootstrap --manifests examples --environment-id UNIQUE_ID
INFO[0000] Reading cluster manifest from examples/cluster.yaml
INFO[0000] Reading master machine manifest from examples/master-machine.yaml
INFO[0000] Reading master user data manifest from examples/master-userdata.yaml
INFO[0000] Creating master machine                      
DEBU[0000] Describing AMI based on filters               bootstrap=create-master-machine machine=test/jchaloup-cama-aws-actuator-testing-machine-master
DEBU[0007] Describing security groups based on filters   bootstrap=create-master-machine machine=test/jchaloup-cama-aws-actuator-testing-machine-master
DEBU[0007] Describing subnets based on filters           bootstrap=create-master-machine machine=test/jchaloup-cama-aws-actuator-testing-machine-master
WARN[0008] More than one subnet id returned, only first one will be used  bootstrap=create-master-machine machine=test/jchaloup-cama-aws-actuator-testing-machine-master
INFO[0009] Master machine created with ipv4: 10.0.102.217, InstanceId: i-0eea29823ae5d50e8
INFO[0014] Waiting for master machine PublicDNS         
DEBU[0014] checking if machine exists                    bootstrap=create-master-machine machine=test/jchaloup-cama-aws-actuator-testing-machine-master
INFO[0014] PublicDnsName: ec2-34-207-227-3.compute-1.amazonaws.com

INFO[0019] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0150] Unable to pull kubeconfig: exit status 255, ssh: connect to host ec2-34-207-227-3.compute-1.amazonaws.com port 22: Connection timed out

INFO[0154] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0158] Unable to pull kubeconfig: exit status 1, Warning: Permanently added 'ec2-34-207-227-3.compute-1.amazonaws.com,34.207.227.3' (ECDSA) to the list of known hosts.
cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0159] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0162] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0164] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0167] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0169] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0172] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0174] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0177] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0179] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0182] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0184] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0187] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0189] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0191] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0194] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0197] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0199] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0202] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0204] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0207] Unable to pull kubeconfig: exit status 1, cat: /etc/kubernetes/admin.conf: No such file or directory

INFO[0209] Pulling kubeconfig from ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0212] Running kubectl --kubeconfig=kubeconfig config set-cluster kubernetes --server=https://ec2-34-207-227-3.compute-1.amazonaws.com:8443
INFO[0217] Waiting for all nodes to come up             
INFO[0222] Waiting for all nodes to come up             
INFO[0227] Waiting for all nodes to come up             
INFO[0232] Waiting for all nodes to come up             
INFO[0237] Waiting for all nodes to come up             
INFO[0242] Waiting for all nodes to come up             
INFO[0248] Is node "ip-10-0-102-217.ec2.internal" ready?: true

INFO[0248] Deploying cluster-api stack                  
INFO[0248] Deploying aws credentials                    
INFO[0248] Creating "test" namespace...                 
INFO[0248] Creating "test/aws-credentials-secret" secret...
INFO[0254] Deploying cluster-api server                 
INFO[0271] Deploying cluster-api controllers            
INFO[0277] Deploying cluster resource                   
INFO[0277] Creating "test/tb-asg-35" cluster...         
INFO[0277] Unable to deploy cluster manifest: unable to create cluster: an error on the server ("service unavailable") has prevented the request from succeeding (post clusters.cluster.k8s.io)
INFO[0282] Deploying cluster resource                   
INFO[0282] Creating "test/tb-asg-35" cluster...         
INFO[0282] Unable to deploy cluster manifest: unable to create cluster: an error on the server ("service unavailable") has prevented the request from succeeding (post clusters.cluster.k8s.io)
INFO[0287] Deploying cluster resource                   
INFO[0287] Creating "test/tb-asg-35" cluster...         
INFO[0287] Unable to deploy cluster manifest: unable to create cluster: an error on the server ("service unavailable") has prevented the request from succeeding (post clusters.cluster.k8s.io)
INFO[0292] Deploying cluster resource                   
INFO[0292] Creating "test/tb-asg-35" cluster...         
INFO[0292] Unable to deploy cluster manifest: unable to create cluster: an error on the server ("service unavailable") has prevented the request from succeeding (post clusters.cluster.k8s.io)
INFO[0297] Deploying cluster resource                   
INFO[0297] Creating "test/tb-asg-35" cluster...         
INFO[0297] Unable to deploy cluster manifest: unable to create cluster: an error on the server ("service unavailable") has prevented the request from succeeding (post clusters.cluster.k8s.io)
INFO[0302] Deploying cluster resource                   
INFO[0302] Creating "test/tb-asg-35" cluster...         
INFO[0302] Reading worker user data manifest from examples/worker-userdata.yaml
INFO[0302] Generating worker machine set user data for master listening at 10.0.102.217
INFO[0302] Creating "test/aws-actuator-node-user-data-secret" secret...
INFO[0302] Reading worker machine manifest from examples/worker-machineset.yaml
INFO[0307] Deploying worker machineset                  
INFO[0307] Creating "test/jchaloup-cama-default-worker-machineset" machineset...