IAAS - Enterprise REST API Integration – ServiceNow & Micro Focus CSA

As an Infrastructure Automation Engineer at La Banque Postale, I designed and developed REST APIs to integrate automated Infrastructure-as-a-Service (IaaS) offerings directly into ServiceNow.

Our infrastructure automation platform, built with Micro Focus Cloud Service Automation (CSA), already supported server provisioning, decommissioning, and resource upgrades. However, we needed to make these services accessible through the company's standard ITSM interface.

I developed the API integration and worked closely with the ServiceNow team to implement approval workflows, authentication, and asynchronous request tracking. Because infrastructure operations could take between 5 and 90 minutes, I implemented an asynchronous architecture allowing ServiceNow to initiate operations and monitor their execution without waiting for an immediate response.

The solution also handled infrastructure provisioning failures, including resource exhaustion in IP address pools or SAN storage pools, ensuring that request statuses remained synchronized between both platforms.

Project Context - Exposing Infrastructure Automation Through REST APIs

At La Banque Postale, our infrastructure automation team was developing self-service offerings for managing enterprise servers. These services were implemented using Micro Focus CSA and supported several infrastructure lifecycle operations:

• Server Provisioning: Create new servers using predefined infrastructure configurations.
• Capacity Upgrades: Increase CPU and memory resources on existing servers.
• Server Decommissioning: Automate server retirement and associated infrastructure operations.

The objective was to make these services directly accessible through ServiceNow, allowing users to request infrastructure changes without interacting with the underlying automation platform. I developed the REST API integration connecting ServiceNow with Micro Focus CSA. This created a standardized interface between the ITSM service catalog and our infrastructure automation workflows.

2. Designing an Asynchronous API Architecture

One of the main technical challenges was handling long-running infrastructure operations. While some operations completed in approximately five minutes, others could take up to 90 minutes, depending on the requested infrastructure and the number of configuration steps.

A synchronous API would have required ServiceNow to wait for the entire operation to complete, creating unnecessary timeout risks and unreliable integrations. I therefore implemented an asynchronous request-processing approach. The workflow was designed as follows:

1. A user submits an infrastructure request through ServiceNow.
2. ServiceNow triggers the corresponding REST API to initiate the automation.
3. Micro Focus CSA starts the infrastructure workflow asynchronously.
4. ServiceNow uses dedicated API operations to retrieve the execution status.
5. When the workflow finishes, ServiceNow updates the request with its final status.

This approach allowed infrastructure operations to execute independently while keeping users informed of their progress.

3. Implementing Security and Approval Workflows

Not every infrastructure request could be executed without validation. Some configurations required significantly more resources than standard server profiles and therefore needed additional approval.

I collaborated with the ServiceNow team to implement approval workflows based on the requested capacity. Requests exceeding predefined resource thresholds were routed through the appropriate validation process before execution. I also implemented a dedicated service account for API authentication, restricting access to authorized ServiceNow integration requests.

This ensured that infrastructure automation remained controlled through the company's established ITSM processes.

4. Handling Infrastructure Failures and Request Synchronization

Infrastructure provisioning doesn't always fail because of an application or automation defect. Some failures are caused by external infrastructure constraints, such as:

• Exhausted IP address pools.
• Insufficient available SAN storage resources.
• Unavailable infrastructure capacity.
• Dependencies that prevent the requested configuration from being completed.

These scenarios reinforced the importance of asynchronous request tracking. I implemented API operations allowing ServiceNow to retrieve execution results and distinguish completed requests from failed operations.

This ensured that users could receive the final outcome of their infrastructure requests, even when execution failed because of resource availability or other infrastructure constraints.

• REST API Design & Development: Designed and implemented APIs exposing Micro Focus CSA infrastructure services for server provisioning, resource upgrades, and decommissioning.
• Asynchronous Workflow Architecture: Implemented asynchronous request execution and status retrieval to support infrastructure operations lasting between 5 and 90 minutes.
• Security & Approval Integration: Configured dedicated service-account authentication and collaborated with the ServiceNow team to implement capacity-based approval workflows.
• Error Handling & System Integration: Developed request-status synchronization and failure reporting for scenarios involving infrastructure resource exhaustion and provisioning errors.

• Integrated 3 infrastructure lifecycle operations for 10 different infrastructure products into ServiceNow: Exposed automated server provisioning, CPU/memory upgrades, and decommissioning through the company's standard ITSM interface.
• Supported operations lasting up to 90 minutes: Implemented asynchronous execution and status tracking, eliminating the need for long-running synchronous API requests.
• Introduced controlled infrastructure provisioning: Integrated capacity-based approval workflows and dedicated service-account authentication to restrict and validate infrastructure changes.
• Improved failure visibility and request tracking: Enabled ServiceNow to retrieve execution outcomes, including failures caused by exhausted IP address pools or unavailable SAN storage resources.