Skip to content

Latest commit

 

History

History
89 lines (68 loc) · 3.66 KB

install_kubernetes.md

File metadata and controls

89 lines (68 loc) · 3.66 KB
Raw

Aptomi Install / Kubernetes Mode

  • You must have your k8s cluster(s) ready to go
  • Aptomi server will be installed in k8s via Helm chart
  • Aptomi client will be installed locally

Installing & Configuring Helm

Our recommendation is to use Helm 2.6.2:

curl https://raw.githubusercontent.com/kubernetes/helm/master/scripts/get | bash /dev/stdin -v v2.6.2
kubectl -n kube-system create sa tiller
kubectl create clusterrolebinding tiller --clusterrole cluster-admin --serviceaccount=kube-system:tiller
helm init --service-account tiller
helm version                                                                                                      17:26:25  ☁  master ☂ ⚡
helm repo add aptomi http://aptomi.io/charts
helm repo update

Installing & Configuring Aptomi

Figure out which context you want to install the Aptomi server to (use kubectl config get-contexts). Then replace [CONTEXT_NAME] with the actual name and run:

helm install --kube-context [CONTEXT_NAME] --name aptomi --namespace aptomi aptomi/aptomi --set users.admin.enabled=true,users.example.enabled=true

To check your deployment progress, run:

watch -n1 -d -- kubectl -n aptomi get pods

Once the Aptomi server is deployed, it will tell you the port it listens on, e.g. 31077:

==> v1/Service
NAME           CLUSTER-IP    EXTERNAL-IP  PORT(S)          AGE
aptomi-aptomi  10.31.252.87  <nodes>      27866:31077/TCP  1s

You can also get an external IP by running kubectl get nodes -o wide, e.g. 35.227.164.206:

NAME                                      STATUS    ROLES     AGE       VERSION        EXTERNAL-IP      OS-IMAGE                             KERNEL-VERSION   CONTAINER-RUNTIME
gke-demo-gke-default-pool-42a191c8-whwg   Ready     <none>    13m       v1.8.7-gke.0   35.227.164.206   Container-Optimized OS from Google   4.4.86+          docker://17.3.2

Install the Aptomi client locally:

curl https://raw.githubusercontent.com/Aptomi/aptomi/master/scripts/aptomi_install.sh | bash /dev/stdin --client-only

Configure the Aptomi client to use the deployed Aptomi server and test it:

vi ~/.aptomi/config.yaml

    api:
      host: 35.227.164.206  <- replace
      port: 31077    <- replace
      
aptomictl version
aptomictl policy show

Accessing the UI

Open the UI at http://APTOMI_SERVER_IP:PORT/ and log in as 'admin/admin'. This is a pre-configured Aptomi domain admin user with full access rights. Once you get going and set up more admin users, you can disable this account or change its password later on.

At this point, most of the UI screens will be empty. This is expected, as Aptomi has no applications imported yet.

Next Steps

You are now ready to point Aptomi to your k8s cluster(s) and start deploying your apps!

Important note before you close this page

You will find in subsequent instructions how to point Aptomi to your k8s cluster(s).

If the Aptomi server itself is running inside your k8s cluster, it may not be able to communicate to the same k8s cluster via external ip:port from the inside.

So, when generating cluster YAMLs and importing them into Aptomi, you will need to tell it to use a local cluster instead of pointing to a specific kubectl context. Luckily, there is a corresponding CLI flag -l, so every time you see aptomictl gen cluster in the instructions:

  • don't forget to use aptomictl gen cluster -l ...
  • instead of aptomictl gen cluster -c [CONTEXT_NAME]. ..

Useful Commands

Upgrade the Aptomi server from latest release -> master

helm upgrade aptomi aptomi/aptomi --reuse-values --set image.tag=master

Restart the Aptomi server

helm upgrade aptomi aptomi/aptomi --reuse-values --recreate-pods