Set up a 3 node HA rancher cluster.
This will create and prep nodes for RKE. This uses the default vpc and subnets.
We create:
- SSH Key Pair
- Security Groups
- 3 Instances
- ELB for 80/443 points to the 3 instances.
Set up a 3 node HA rancher cluster.
This will create and prep nodes for RKE. This uses the default vpc and subnets.
We create:
docker rm -fv $(docker ps -qa) | |
docker volume ls -q | xargs docker volume rm | |
rm -rf /etc/ceph \ | |
/etc/cni \ | |
/etc/kubernetes \ | |
/opt/cni \ | |
/opt/rke \ | |
/run/secrets/kubernetes.io \ | |
/run/calico \ | |
/run/flannel \ |
storageClass with PersistentVolumeClaims are really the way to do storage with Kubernetes.
For Azure There are 3 types of storage avalible.
First and foremost it is important to not run out of connections from the Rancher Server to the MySQL database server. The threads per server are driven by the thread pool size. So max connections should follow this basic formula:
Number of nodes x Total threads + buffer for non-rancher tasks
Create an cluster with an existing vnet/subnet in your subscription
You need to set up an app access (service principal): https://rancher.com/docs/rancher/v1.6/en/hosts/azure/#app-registration
apiVersion: extensions/v1beta1 | |
kind: Deployment | |
metadata: | |
creationTimestamp: null | |
labels: | |
run: server | |
name: server | |
spec: | |
replicas: 1 | |
selector: |
#!/bin/bash | |
### kubectl should really allow you to export a singel context. | |
### if you only have one context you can just run `kubectl config view --flatten` | |
### if you have more than one context use this script. | |
### | |
### **Requires jq: https://stedolan.github.io/jq/ | |
### | |
context=${1:-"minikube"} |