1
0
Fork 0
molgenis-ops-docker-helm/README.md

140 lines
4.3 KiB
Markdown
Raw Normal View History

2018-05-30 23:21:29 +02:00
# MOLGENIS Helm templates
2018-06-27 21:20:13 +02:00
These are the Helm templates that we will use for MOLGENIS operations. Basic concepts in respect to docker you need to know.
**Deployments**
Are a set of pods that will be deployed according to configuration that is usually managed bij Helm. These pods interact with eachother by being in the same namespace created by kubernetes according to the deployment configuration.
**Pods**
A pod is wrapper around a container. It will recreate the container when it is shutdown for some reason and interact with other pods when needed.
**Containers**
A container is a docker-container that is created from a docker image. It could be seen as an VM for example
**Images**
An image is a template for a container some sort of boot script but also contains the os for example. A build dockerfile, if you will.
**Prerequisites**
There are some prerequisites you need.
2018-05-30 23:21:29 +02:00
2018-06-27 15:17:32 +02:00
- docker
- minikube
2018-06-27 21:20:13 +02:00
## Kubernetes
When you want to use kubernetes there are some commands you need to know. Also running on a remote cluster will be a must have to control your whole DTAP.
### Useful commands
2018-05-30 23:21:29 +02:00
2018-06-27 15:17:32 +02:00
Commands that can be used to get information from a kubernetes cluster
**Pods**
- ```kubectl get pods (optional: [--all-namspaces])```
2018-06-27 21:20:13 +02:00
2018-05-30 23:21:29 +02:00
Gets alls running instances of containers from a certain deployment
- ```kubectl describe pod #pod name# --namespace=#namesspace#```
Describes the pod initialization, also displays error messages more accurately if they occur
- ```kubectl remove pod #pod name# --namespace=#namespace# (optional: [--force] [--grace-period=0])```
Removes a pod from the system (but will restart if the option is set in the deployment,yaml *[see note]*).
2018-06-27 21:20:13 +02:00
**note:** You can not do this while the deployment of the service is still there
**Services**
- ```kubectl get services```
2018-05-30 23:21:29 +02:00
Gets all services from a deployment
**Volumes**
2018-06-27 21:20:13 +02:00
- ```kubectl get pv```
2018-05-30 23:21:29 +02:00
Gets all persistant volumes
2018-06-27 21:20:13 +02:00
- ```kubectl get pvc```
2018-05-30 23:21:29 +02:00
Gets all persistent volume claims
**Deployments**
2018-06-27 21:20:13 +02:00
- ```kubectl get deployments```
2018-05-30 23:21:29 +02:00
Gets all deployments (comparable with docker-compose)
2018-05-30 23:21:29 +02:00
2018-06-27 21:20:13 +02:00
## Remote clusters
2018-06-27 15:17:32 +02:00
When you want to see what is running on the clusters at the CIT you have to make a context switch.
You can access the cluster with kubeconfig-files. You can obtain these by downloading them from the
MOLGENIS kubernetes cluster.
- Go to https://rancher.molgenis.org:7777 and login
2018-06-27 21:20:13 +02:00
- Go to Rancher --> Cluster: *#name#* --> *Kubeconfig File*
- Go to a **Terminal** where ```kubectl``` is available
- Add this configuration to ~/.kube/config (or place a new file besides this one)
*Example*:
```bash
# When you added the MOLGENIS configuration to the original configuration
kubectl config use-context molgenis
2018-06-27 21:20:13 +02:00
# or when you placed the MOLGENIS configuration besides the original one
kubectl config use-context molgenis --kubeconfig=*full path to molgenis config*
```
- You can now access all facilities of the MOLGENIS cluster like it is running locally
*Example:*
```bash
kubectl get pods --namespace=*#namespace of application#*
```
## Helm
2018-06-27 22:12:27 +02:00
This repository is serves also as a catalogue for Rancher. We have serveral apps that are served through this repoistory. e.g.
2018-06-27 15:17:32 +02:00
2018-06-27 22:12:27 +02:00
- [Jenkins](molgenis-jenkins/README.md)
- [NEXUS](molgenis-nexus/README.md)
- [HTTPD](molgenis-httpd/README.md)
2018-07-14 13:13:23 +02:00
- [MOLNIGES preview](molgenis-preview/README.md)
2018-06-27 15:17:32 +02:00
2018-06-27 22:12:27 +02:00
### Useful commands
You can you need to know to easily develop and deploy helm-charts
- ```helm lint .```
To test your helm chart for code errors.
- ```helm install . --dry-run --debug```
Check if your configuration deploys on a kubernetes cluster and check the configuration
2018-05-30 23:21:29 +02:00
2018-07-14 13:13:23 +02:00
- ```helm install . #release name# --namespace #remote namespace#```
2018-06-27 21:20:13 +02:00
2018-05-30 23:21:29 +02:00
Do it in the root of the project where the Chart.yaml is located
It installs a release of a kubernetes stack. You also store this as an artifact in a kubernetes repository
2018-06-27 21:20:13 +02:00
- ```helm list```
2018-05-30 23:21:29 +02:00
Lists all installed releases
- ```helm delete #release#```
2018-06-27 21:20:13 +02:00
2018-06-27 22:12:27 +02:00
Performs a sort of mvn clean on your workspace. Very handy for zombie persistent volumes or claims.
2018-07-14 13:13:23 +02:00
- ```install tiller on remote cluster```
To install tiller on a remote cluster you need an rbac-config.yml.
```kubectl create -f rbac-config.yaml```
When you have defined the yaml you can add the tiller to the cluster by following the steps below.
```helm init --service-account tiller```