After following steps in guide to create certs:
$ cp ca.cert.pem ~/.helm/ca.pem
$ cp helm.cert.pem ~/.helm/cert.pem
$ cp helm.key.pem ~/.helm/key.pem
Then use the –tls options:
$ helm ls --tls
Kubernetes: can’t schedule a pod on master node
Attempting to run a pod on the master and get this error:
Warning FailedScheduling 14m (x2 over 14m) default-scheduler 0/1 nodes are available: 1 node(s) had taints that the pod didn't tolerate.
From issue here: you can configure the master node to run pods with:
kubectl taint nodes --all node-role.kubernetes.io/master-
Running GitLab in a Docker container on a different port
Gitlab by default runs on port 80. GitLab in a Docker container runs the same as a when natively installed, but to change the port you need to change the config, and change the exposed ports on the container.
First, per steps here, start the container with:
docker run --detach \
--hostname gitlab.example.com \
--publish host-https-port:container-https-port
--publish host-http-port:container-http-port
--publish 22:22 \
--name gitlab \
--restart always \
--volume /srv/gitlab/config:/etc/gitlab \
--volume /srv/gitlab/logs:/var/log/gitlab \
--volume /srv/gitlab/data:/var/opt/gitlab \
gitlab/gitlab-ce:latest
By default host-http-port and container-http-port are 80, and host-https-port container-https-port are 443. Change these to be whatever port you want to run on, but keep each pair the same (e.g. host-http-port and container-http-port = 8090)
When the container is up, start a shell into the running container:
docker exec -it containerid sh
and then edit the config file:
vi /etc/gitlab/gitlab.rb
and add line (at the top is ok):
external_url 'http://localhost:your-new-port-here
setting your-new-port-here to the new port.
Reconfigure the server:
gitlab-ctl reconfigure
gitlab-ctl restart
Done!
Rancher RKE Kubernetes install notes
Rancher’s RKE is a Kubernetes cluster installer – see more here.
Pre-reqs:
- Docker must be running on the client machine where you are going to run the rke setup tool
- The docs are not obvious, but the rke tool is run on a client machine to provision your cluster, it is not run on any of the target cluster nodes
Notes using Ubuntu 16.04 server.
Remove prior Docker installs:
sudo apt-get remove docker docker-engine docker.ioCreate docker group and add user to docker group:
sudo groupadd docker
usermod -aG docker <user_name>
Install per Docker CE install steps here, or use the Rancher provider install script here.
Supported Docker versions for RKE (as of Dec 2018) are: 1.11.x 1.12.x 1.13.x 17.03.x
Configure Docker daemon to listen for incoming requests on 2376, as per steps here.
Using ‘rke config’ with the default/minimal cluster.yml here, and then install/setup with ‘rke up’
If you didn’t change the name of the cluster.yml file, after the install is complete, you’ll have a kube_config_cluster.yml file in the same dir which you can use with kubectl to interact with you cluster, or add it into your existing ~/.kube/config file