Note: This work currently has significant requirements on k8s storage, VolumeGroup (aka consistency groups), which are currently a KEP.

• KEP-3476: Add Volume Group KEP{}

The deliver of this KEP will significantly impact the delivery of this feature.

Goal

To protect VM workload with a low Recovery Time Objective (RTO of minutes) and Recovery Point Objective (RPO of several minutes). This is required for standard IT workloads, as well as high performance databases. This epic is focused on VMs using storage deployed on OpenShift Data Foundation (ODF) and orchestrated via ACM, using asynchronous replication of snapshots. 

User Stories

• As a VM workload administrator/owner, I want to protect a VM based application so I can recover it at another site, with minimal data loss, and minimal recovery time.
  • Basic VM, one disk, no application workload required.
  • Windows and Linux
• As a database administrator/owner, I want to protect a my database based application so I can recover it at another site, with minimal data loss, and minimal recovery time.
  • Basic VM with database workload
    • HammerDB or something similar workload generator.
  • Windows and Linux
  • Success criteria
    • VM restart on DR site
    • Database recovers
    • Workload continues from the last committed transaction from the last complete replication before the failure. This WILL involve data loss, as expected.

Non-Requirements

• Performance and scale

Notes

• Any additional details or decisions made/needed

Done Checklist