Ansible Automation for VMware Cloud Foundation Deployment with Hitachi VSP External Storage
As enterprise IT environments grow increasingly complex, the demand for infrastructure solutions that combine speed, resilience, and operational simplicity has never been greater. Hitachi VSP One has pivotal capabilities that can be leveraged to deploy VMware Cloud Foundation 9, with its native support for Fibre Channel-based VMFS storage — enabling organizations with established Hitachi VSP One to seamlessly adopt a modern cloud management framework without compromising their existing storage architecture.
This blog documents how Hitachi’s Integrated Platforms Solutions Engineering team tackled end-to-end VCF deployment by building a unified Ansible automation framework that orchestrates Brocade FC switching, Hitachi VSP One Block storage provisioning, and VCF 9.0.1 deployment in a single, repeatable pipeline — delivering a 79% reduction in deployment time and an operational model that is consistent, auditable, and ready to scale.
The Purpose Behind the Pipeline
Behind every infrastructure deployment is a team of specialists — a SAN admin, a storage engineer, a VMware architect — each working in their own silo, handing off to the next only after their piece is complete. When something breaks, nobody owns the full picture. When a deployment must be repeated, everyone starts from scratch.
The core conviction driving this work is simple: infrastructure knowledge should not live in people's heads. It should live in code — in version-controlled playbooks and inventory files that every team member can trust, every auditor can inspect, and every customer can benefit from consistently.
Infrastructure Architecture Overview
The automation framework is designed as three Ansible role collections — one per infrastructure layer — executed in sequence from a single control node. All credentials are Ansible Vault-encrypted. All environment-specific variables (WWNs, IPs, LDEV IDs, cluster names) live in inventory files completely separated from playbook logic.
Using the hitachi.vspone_block certified Ansible collection (Red Hat
Before any host can see storage, zone aliases must be created for every ESXi HBA port and every VSP target port on the Brocade switch. Zones are then built in a target-based naming model, and the zone configuration is activated — all for both SAN fabrics in a single playbook run. A critical safety feature: the brocade.fos modules use transactional rollback. If any task fails mid-run, the zoning database reverts to its prior clean state, preventing partial configurations.
VMware vSphere import to VCF 9.0.1
VCF 9 introduced a major architectural shift: VCF Operations is now the primary management interface, with SDDC Manager as a component within it. Critically, VCF 9 also introduced native FC-based VMFS as principal storage for management and workload domains — making Hitachi VSP integration a first-class deployment path, not a workaround.
The automation layer deploys the vCenter Server OVA, configures the datacenter, cluster HA and DRS, deploys NSX Manager, and calls the VCF Operations API to import the environment as a workload domain — all in a single unattended execution.
-
FC Zoning (brocade-zoning playbook): Zone aliases are created, zones are built per the naming convention in inventory, and the configuration is activated across both SAN fabrics. ~8 min.
-
Storage Provisioning (storage_provisioning playbook): ESXi hosts are registered on the VSP array, LDEVs are created, and host groups are mapped per FC port. The array is now presenting storage to the hosts. ~22 min.
-
HBA Discovery (scan_hba playbook): The first playbook collects HBA WWNs from ESXi hosts via the vSphere API and writes them back to host_vars files — eliminating manual WWN lookups from switch nameservers or array management tools. ~5 min.
-
Datastore Creation (datastore_create playbook): vCenter scans for newly visible LUNs and formats them as VMFS datastores with storage policies applied. ~7 min.
-
VCF Deployment (vcsa_deploy playbook): vCenter Server OVA is deployed, cluster objects configured, NSX Manager deployed and registered, and the VCF import workflow triggered via VCF Operations. ~40 min.
Measured Results: 79% Time Reduction
Every task was measured across two real-world deployment scenarios. The numbers below are actual — not projected — based on deployments tracked by the Integrated Platforms Solutions Engineering team.
The existing vSphere import result at 85.9% reduction (24 minutes vs. 170 minutes ) is especially significant for customers with an existing Hitachi VSP investment. Because storage and zoning are already in place, the automation handles only incremental provisioning and the VCF import — with no domain expert required to be online during execution.
Hitachi VSP One Block and Ansible provides a robust, proven blueprint for building resilient, scalable, and efficiently managed hybrid cloud environments deployed using VCF 9.0. By combining the “hitachi.vspone_block” certified collection for storage provisioning, the “brocade.fos” collection for FC fabric automation, and VMware Ansible modules for VCF deployment, enterprises can modernise their data centres while achieving continuous operational excellence.
With native FC-VMFS support in VCF 9, customers with existing or new Hitachi VSP One infrastructure are positioned to adopt VCF without rebuilding their storage architecture. The framework is extensible to additional VSP models, additional workload domains, and full Day-2 lifecycle operations.
References & Resources