A client has 2 sites, each with 3 virtual nodes; the primary site is in primary data center, and the secondary site is in the DR data center.
The client now wants to migrate both sites to physical appliances whilst minimizing cluster downtime.
Assume that all other cluster pre-reqs will be met, such as: software+patch levels, licenses, SSL certificates, etc, would this high level procedure be the best way to accomplish that task (migrate while minimizing downtime):
- In the DR site, add physical devices to the existing secondary site
- this can be done whilst the cluster is online - if I recall correclty?
- Configure a NEW Secondary site in the primary datacenter with the physical devices
- this will require cluster downtime to configure
- Start the cluster to allow replication to the new secondary site.
- After the NEW Secondary site has fully replicated, take down the cluster again
- Promote the NEW site to primary (Physical)
- Remove old Primary Site (VMs)
- Remove virtual nodes from secondary site configuration
- Start the cluster
- Update DNS entries to point to new primary site VIP and secondary site VIP, accordingly
am i forgetting a major step?
is there a better way?
are there any pitfalls with this approach?
is this documented somewhere?