This article is contunuation of the previouse one: VCF Fleet Basic with Multiple VCF Instances
Next, let’s move from basic fleet management to site‑level high availability within a region.
This design uses what you may already know as a stretched cluster:
-
We have a single VCF instance spanning two availability zones.
-
Compute and storage are stretched between the zones, giving us continuous availability and IP portability across them.
The goals of this design are:
-
To implement site high availability within a region.
-
To protect against the failure of a fault domain, such as an entire availability zone.
-
To distribute resources across fault domains so that an outage in one zone has minimal impact.
Because IPs are portable across the zones, applications can continue running even if one zone fails, and your failover can be largely transparent.
With this kind of design, details like:
-
Where we place NSX Edges,
-
How we handle north‑south traffic, and
-
How we manage egress independently from each side or in a shared pattern
become very important. There are several valid ways to do this, and which one you choose depends on customer requirements.
Please check next article as logical continuation: VCF Fleet with Disaster Recovery (Across Regions)
