The diagram path now follows a documented standard procedure (lookup the closest Oracle Architecture Center reference → confirm components → author absolute_layout → spec validator → render → visually verify) and ships persistent guardrails so layout regressions can't recur. Persistent procedure changes (apply to all users, all sessions): - tools/diagram_spec_validator.py — geometry checks (CONTAINER_TOO_THIN, CONTAINER_PADDING_VIOLATION, LABEL_OVERFLOW_PARENT) run BEFORE either renderer (drawio + PPTX). Catches the subnet-collapse / label-overflow bugs that the post-render drawio validator missed. - tools/oci_diagram_gen.py + tools/oci_pptx_diagram_gen.py — call the spec validator before emitting any output. Adds mysql / mysql_heatwave type aliases. - tools/archcenter_pattern_lookup.py — scores against cached page descriptions (not just the 1-line summary), supports --queries for multi-fragment composition, and applies synonym expansion via kb/architecture-center/synonyms.yaml so "LB HA cross AD" matches "load balancer high availability availability domain". - kb/architecture-center/synonyms.yaml — canonical synonym table (load balancer, autonomous database, data guard, …) used by the lookup scorer. KB enrichment: - tools/archcenter_description_fetcher.py + 121 cached _description.md under kb/diagram/assets/archcenter-refs/<slug>/. Removes the runtime dependency on docs.oracle.com when authoring specs and feeds the pattern-lookup scorer. - 110+ cached .drawio / .svg / .png references for offline reuse, plus the OCI Toolkit v24.2 import (kb/diagram/assets/oci-toolkit-drawio). Documentation: - docs/skill/output-formats.md — new "Standard diagram-generation procedure (MANDATORY)" + geometry rules + the new validator entry. - SKILL.md option 2 — references the mandatory procedure. - README.md — describes the spec validator, archcenter_pattern_lookup and description fetcher, and updates the KB-health table. Tooling that backs the procedure (cumulative across recent sessions): tools/archcenter_case_runner.py, archcenter_batch_driver.py, archcenter_zip_downloader.py, drawio_visual_validator.py, drawio_fidelity_eval.py, harvest_drawio_icon.py, import_oci_library.py, oci_pptx_diagram_gen.py, oci_pptx_render.py, refresh_pptx_icon_index.py. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
7.6 KiB
Deploy a VMware vSAN Stretched Cluster across OCI Regions with Oracle Cloud VMware Solution
- Source: https://docs.oracle.com/en/solutions/ocvs-vsan-stretched-cluster/index.html
- Date: 2025-07
- Type: reference-architecture
- Services: ocvs
- Tags: vmware, ha-dr
Summary (catalog)
vSAN Stretched Cluster across OCI regions for VMware HA/DR. Synchronous storage replication between sites, automatic VM restart on site failure. Witness host in third location.
Architecture (fetched from source)
Architecture
This architecture shows how you can deploy custom VMware vSAN stretched clusters across multiple OCI regions.
The high-level topology consists of:
-
Primary Site : Oracle Cloud VMware Solution SDDC deployed in OCI Dedicated Region A.
-
Secondary Site : Oracle Cloud VMware Solution SDDC deployed in OCI Dedicated Region B.
-
Witness Site : A regionally separate location for deploying the VMware vSAN Witness Appliance.
Communication across these sites is established through OCI’s private backbone and OCI FastConnect , both of which are mandatory to meet the low-latency and high-bandwidth requirements of a stable VMware vSAN stretched cluster.
Note: IPSec VPN is not supported for this configuration.
The following diagram illustrates this architecture.
Description of the illustration ocvs-vsan-stretched-cluster.png
ocvs-vsan-stretched-cluster-oracle.zip
The following sections outline the key technical considerations that influence a successful deployment of a VMware vSAN stretched cluster in Oracle Cloud VMware Solution across OCI Dedicated Region s.
Networking Considerations
A key enabler of this architecture is the robust OCI backbone network that interconnects OCI Dedicated Region s within a customer tenancy. This backbone ensures the high-speed, low-latency communication necessary for VMware vSAN replication traffic and heartbeat signaling between sites.
Key factors to plan for:
-
Establish Remote Peering Connections (RPCs) between the VCNs in OCI Dedicated Region A and OCI Dedicated Region B using Dynamic Routing Gateways (DRGs) . This enables full mesh connectivity across all VMware ESXi hosts.
-
Use OCI FastConnect (not IPSec VPN) to connect both OCI Dedicated Region s to the public OCI region hosting the Witness. This ensures consistent low-latency and reliable throughput to support witness communication.
-
Reference documentation: Remote Peering , Managing DRGs , OCI FastConnect
Compute and Storage Considerations
Infrastructure planning across all three regions involves several decisions:
-
Region Selection
-
Choose two OCI Dedicated Region s with < 5 ms RTT latency between them.
-
Select a public OCI region with < 200 ms RTT latency to both OCI Dedicated Region s for Witness deployment.
-
Shape Selection
-
Use Dense Bare Metal shapes (e.g., BM.DenseIO.E5.128) with local NVMe storage for VMware vSAN.
-
Avoid Standard shapes that use Block Volumes, as they are not suitable for stretched vSAN deployments.
-
Minimum Host Requirements
-
Primary Region : Minimum three Dense Bare Metal hosts
-
Secondary Region : Minimum three Dense Bare Metal hosts
-
Witness Region : One Bare Metal host
-
Witness Appliance Guidelines
-
Follow vSAN Witness Design guidance .
-
Always refer to the official documentation from Broadcom to get the latest updates as the requirements could change. Below are some references:
-
Stretched cluster considerations in VCF 5.1.2
-
Minimum host count for vSAN stretched clusters
Stretched Cluster Requirements
-
RTT latency < 5 ms between Primary and Secondary Regions
-
RTT latency < 200 ms between either site and the Witness node
-
All hosts (including Witness) must belong to the same VMware vSAN cluster
-
Host hardware and configuration must be identical across regions
-
Witness must reside in a third, separate location
Operational Considerations
Customers are responsible for completing Day 2 operations manually. Key notes:
-
Oracle Cloud VMware Solution environments are deployed separately in each OCI Dedicated Region . The secondary site’s VMware vCenter and VMware NSX Manager must be manually detached and integrated with the primary cluster.
-
Manual failover and route updates are required in case of a site failure.
-
VMware NSX Tier-0 Gateway is active only in one site, implying an active-passive model for North-South traffic routing.
Design Overview
Building on the previous sections that covered architecture and requirements for a stretched vSAN configuration with Oracle Cloud VMware Solution , this section explains how to implement a highly available design capable of withstanding the failure of an OCI Dedicated Region .
This design uses two VCNs per site , resulting in a total of four VCNs :
OCI Dedicated Region A
- VCN Primary with two CIDR blocks; for example, 10.16.0.0/16 as the primary and 172.45.0.0/16 as the secondary CIDR (added after VCN creation). The secondary CIDR is required only for the initial SDDC deployment.
Since an Oracle Cloud VMware Solution SDDC cannot span multiple VCNs, a secondary CIDR block ( 172.45.0.0/16 ) is attached to the Primary VCN within OCI Dedicated Region A. This enables VLAN definitions for management and services subnets while keeping them logically grouped within a single VCN.
- VCN MGMT Active , using the same CIDR block as the secondary CIDR attached to VCN Primary; i.e., 172.45.0.0/16 .
OCI Dedicated Region B
-
VCN Secondary with a CIDR block distinct from and non-overlapping with VCN Primary ; for example, 10.17.0.0/16 .
-
VCN MGMT Failover , using the same CIDR block as VCN MGMT Active ; i.e., 172.45.0.0/16 .
Oracle Cloud VMware Solution provides flexibility in network provisioning. During SDDC creation, users can either:
-
Specify a CIDR block and allow Oracle Cloud VMware Solution automation to create required networking components, or
-
Manually create VCNs, subnets, VLANs, route tables, and NSGs beforehand, then select these existing resources during deployment.
For this stretched vSAN design, the latter approach is necessary. Precise control over network segmentation across multiple VCNs and regions requires pre-creation of route tables, NSGs, and VLANs. This separation supports clear responsibilities between VCNs and enables seamless failover behavior.
A critical aspect is that the management subnet ( 172.45.0.0/16 ) must be accessible in both OCI Dedicated Region s. To support failover, the design allows this VCN MGMT network to “float” between the two sites via manual network updates during failover events, such as modifying route tables and re-advertising the subnet through DRG attachments.
DNS resolution is vital for failover and service availability. Therefore, a dedicated services subnet will be created in each VCN to host DNS and supporting infrastructure.
For VLAN tagging simplicity:
-
VLAN tags in the 100 range are region-specific , confined to their respective sites.
-
VLAN tags in the 200 range are associated with the 172.45.0.0/16 subnet and will float between sites .
With the high-level design defined, we now step into the practical configuration of each site, starting with the Primary region.
Title and Copyright Information
Deploy a VMware vSAN Stretched Cluster across OCI Regions with Oracle Cloud VMware Solution
G36844-01
July 2025
Copyright © 2025,
Oracle and/or its affiliates.