You can find the Camunda Cloud Helm charts here:
https://github.com/camunda/camunda-cloud-helm
If you decide to try this yourself you may run into a couple of issues with Elasticsearch. First, you may get messages from your ES container complaining about operating memory being too low as in:
vm.max_map_count = 262144
Easy enough to fix in Linux. Issue the following command:
sysctl -w vm.max_map_count=262144
or set the value in sysctl.conf.
For Windows users using Docker Desktop with WSL2 you need to create a .wslconfig file in your Users/{user} directory with the entry vm.max_map_count=262144. Restart Desktop and you should be good.
Great, so now Elasticsearch starts but it can’t find a master node and so the elasticsearch cluster never bootstraps. Exporter, Operate, Tasklist fail to work. I borrowed the values.yml file from here:
https://github.com/camunda/camunda-cloud-helm/blob/main/charts/ccsm-helm/values.yaml
And if I understand it correctly, the values.yml file allows you to update and augment parameters when installing. I added the following precisely at line 449 in the "extraEnvs’ stanza:
- name: "cluster.initial_master_nodes"
value: "elasticsearch-master-0"
And if I understand this correctly this is telling Elasticsearch to look to elasticsearch-master-0 for the initial master node. You then need to issue the following command to use the values file:
helm install test camunda-cloud/ccsm-helm -f values.yml
Lo and behold it started working! Issue the kubectl port forwarding commands as instructed by the helm install output (For Zeebe, Operate, and Tasklist) and enjoy! Or you can use ingress instead of port forwarding.