Learn how the VMware Aria Automation 8.18.1 to VCF Automation 9.1 upgrade works under the hood, covering the Fleet Lifecycle architecture, each migration phase, and exactly what happens to your data and services throughout the transition.
Overview
Upgrading from VMware Aria Automation 8.18.1 to VMware Cloud Foundation (VCF) Automation 9.1 is more than a version increment. The process introduces a carefully orchestrated migration architecture designed to safely transition your environment while preserving all existing automation content.
While the official upgrade guide covers each required step, this post explains the architecture behind the upgrade, the purpose of each migration stage, and what happens to your data and services throughout the process. Understanding this context will help you prepare more effectively for the transition.
Environment Details
This upgrade path applies to the following environment:
- Source: VMware Aria Automation 8.18.x (minimum version 8.18.1)
- Target: VCF Automation 9.1
Prerequisites
Ensure your environment meets the following requirements before initiating the upgrade through VCF Operations.
Technical Requirements:
- Import: VMware Aria Automation 8.18.1 or later must be imported into VCF Operations.
- Networking: Allocate a dedicated /29 CIDR block with a range of eight contiguous IP addresses for the cluster. Create a new VCF Services Runtime 9.1 FQDN that resolves to a unique IP address outside the dedicated CIDR range. This FQDN serves as the internal endpoint for the VCF Services runtime instance hosting the VCF Automation 9.1 nodes and is used exclusively for communication between VCF components.
- Infrastructure: Download the VCF Automation 9.1.0.0 upgrade binaries and install the Migration Service Engine binary through the software depot.
The Fleet Lifecycle
The Fleet Lifecycle oversees the comprehensive lifecycle of management components, including end-to-end upgrades, patching, day-2 operations, and backup and recovery processes. It facilitates the transition to version 9.1.x for environments running VMware Aria Automation 8.18.x or VCF Automation 9.0.x.
Video walkthrough: A video overview of the Fleet Lifecycle upgrade process is available for reference.
Upgrade Process: Stage-by-Stage Breakdown
The transition from VMware Aria Automation 8.18.x to VCF Automation 9.1 follows a highly orchestrated workflow. The process divides into an initial import and orchestration phase, followed by a core 16-stage component upgrade.
Phase 1: Import and Orchestration
Before the upgrade begins, the Fleet Lifecycle collects the inputs needed to deploy the target environment. This top-level orchestration wrapper handles the following steps:
- Validation and Deployment: Validates all required configuration parameters and deploys the new VCF 9.1 service runtime and dedicated clusters as needed.
- Stage and Import: Stages the VCF Automation and bundle packages, then imports the existing VMware Aria Automation component to extract necessary details such as the component ID and sizing requirements.
- Pre-checks: Runs the comprehensive five-group pre-check sequence (connectivity, source, health, topology, application) against the target clusters.
- Trigger: Once the component is imported and validated, the Fleet Lifecycle initiates the core upgrade workflow on the cluster.
Phase 2: Preparation and Export (Stages 1 to 8)
In this phase, the source environment prepares for migration:
- Snapshots: Quiesced VM snapshots (without memory) are created through vCenter to provide a safe rollback point.
- Graceful Shutdowns: All pods and services on the source Aria Automation 8.18.x nodes are stopped via SSH.
- Data Export: The Aria Automation 8.18.x databases and configuration data are exported to “Hard disk 2”.
- All remaining internal pods and services are stopped, and the source VM name is stored securely.
| Stage | Task Name | Description |
| 1 | Provide Source Node IP Addresses | Collects all Aria Automation 8.18.x node IP addresses to prepare for remote execution. |
| 2 | Create a Snapshot of the Source Environment | Creates quiesced VM snapshots (without memory) of all source nodes via vCenter. |
| 3 | Stop Source pods and Services | Gracefully stops all pods and services on the source via SSH. |
| 4 | Export Source System Data | Exports databases and configuration data onto “Hard disk 2”. |
| 5 | Stop All Source Services | Stops all remaining internal and background services on the source nodes. |
| 6 | Provide Source Environment IP Address | Retrieves the primary IP address of the source VM. |
| 7 | Identify Source VM Name from IP Address | Resolves the vCenter VM name from the IP address to identify the disk source. |
| 8 | Confirm Source VM Name | Stores the resolved VM name for use in downstream mounting tasks. |
Phase 3: The Downtime Window (Stages 9 to 12)
Downtime begins at Stage 9, when the original Aria Automation 8.18.x environment is powered off entirely. The Fleet Lifecycle then detaches “Hard disk 2” from the source node and mounts it to the target VCF Automation node.
| Stage | Task Name | Description |
| 9 | Shut Down the Source Environment | Triggers OS-level shutdown of all Aria Automation 8.18.x nodes and polls until they are offline. |
| 10 | Attach Source VMDKs to the Target Environment | Mounts “Hard disk 2” from the source VM to the new target VCF Automation node. |
| 11 | Store the Resolved Environment Name | Persists the resolved VM name to durable storage (workaround for race conditions). |
| 12 | Create a Data Migration Indicator | Enables native data restoration from the mounted disk during cluster deployment. |
Phase 4: Post-Deployment and Cleanup (Stages 13 to 16)
Once the new VCF Automation 9.1 cluster deploys successfully and begins serving traffic, the system initiates the post-deployment workflow:
- The system retrieves the original VM details and unmounts “Hard disk 2” from the new deployment, returning it to the original source VM.
- All temporary Kubernetes upgrade secrets are cleaned up to complete the process.
| Stage | Task Name | Description |
| 13 | Restore resolved VM name | Restores the persisted VM name after the new 9.1 VCF Automation cluster is live. |
| 14 | Provide mounted source VM name | Reads the restored VM name and exposes it as a task output for the unmount operation. |
| 15 | Unmount VMDKs from the target environment | Detaches “Hard disk 2” from the new node and returns it to the original source VM. |
| 16 | Clean up upgrade configuration files | Deletes the temporary Kubernetes upgrade configuration secrets. |
Summary
The orchestrated upgrade to VCF Automation 9.1 modernizes your infrastructure with minimal risk by isolating migration data onto a dedicated disk and transferring it cleanly to the new cluster. Each phase is designed to preserve your existing automation content while providing clear rollback options at every stage.
Acknowledgements: Special thanks to Arun Nukula, Iyaz Khan, and Owen Sheehy
Discover more from VMware Cloud Foundation (VCF) Blog
Subscribe to get the latest posts sent to your email.
