Goal:

• Mount etcd in a data volume (Data Disk) and evaluate performance running cluster workloads

Motivation:

The OPCT reported that Azure does not complain with the check rules when running the provider benchmark activity (OPCT-264). Exploring the details we can see the culprit is the etcd is not performing when under workloads, mainly when pressure the OS Disk - where etcd is mounted sharing with OS and many other control plane workloads.

The resource isolation for sensitive workloads is recommended widely for providers[1], community[2], and Red Hat[3], considering OCP clusters in Azure (baseline IPI) has allocated a considered amount of compute and storage resources and still reporting issues, that isolate and analysis became more important.

[1]  https://learn.microsoft.com/en-us/azure/architecture/best-practices/data-partitioning

[2]  " As a more long term solution, I would like to look into creating a data disk for etcd." 

• https://github.com/kubernetes-sigs/cluster-api-provider-azure/issues/448
   
  [3]  "you should set up storage in any manner that best fits your needs using standard Linux commands, but we recommend that you mount a large device on /var/lib/containers."

• https://www.redhat.com/en/blog/working-container-storage-library-and-tools-red-hat-enterprise-linux