Kind cluster setup - pods failing

[Mahendra_Gondi]

I am trying to setup kind cluster on my MacBook, followed comunda 8 docs and changed helm memory configurations to make sure it uses 64mb for kind cluster setup and ran all commands successfully.

elastic search pods are not starting which causes Zeebe,operate and task list also failing.

Elasticsearch container log shows “Defaulted container elasticsearch out of: elasticsearch, sysctl (init)”

any help would be much appreciated

[Alex_Voloshyn]

Hi @Mahendra_Gondi

1. Can you share the helm chart properties file that you use?
2. Output of: kubectl get storageclass
3. Output of: kubectl describe pvc -n <your_namespace> <elasticsearch_master_pod_name>

Regards,
Alex

[Ingo_Richtsmeier]

Hi @Mahendra_Gondi,

I remember I had a similar issue that pods could not start due to a lack of main memory.

After increasing the memory limit in my Docker desktop settings to 13.5 GB, the cluster was able to start successfully.

Hope this helps, Ingo

[nathan.loding]

Hi @Mahendra_Gondi - I also had to give Docker more memory on my Macbook in order to get kind running properly. Kind is “Kubernetes in Docker”, so essentially it’s running all the Camunda images inside another image. It’s not the most efficient resource usage, but it’s very convenient for local development and testing because it’s easy to use.

Try increasing the amount of resources available to Docker. (The screenshot below is my current Docker configuration; yours may not need the same resources, I am providing it just as an example.) And if that doesn’t work, please share a more complete log as well as the events from the pod (see “Gathering Events and Logs” in this blog post).