Digital Realty
Digital Realty is one of the world’s largest providers of data center, colocation, and interconnection services, with facilities spanning several continents and an enterprise customer base that includes tech giants like AWS, Oracle, IBM, and Microsoft Azure.
Case Details
Client: Digital Realty
Tags: Data Centers, DCIM
Project Duration: 4 years
Client website: digitalrealty.com
Let’s Work Together
Call us directly, submit a sample or email us!
The Story
Digital Realty is one of the world’s largest providers of data center, colocation, and interconnection services, with facilities spanning several continents and an enterprise customer base that includes tech giants like AWS, Oracle, IBM, and Microsoft Azure. To manage and monitor this massive infrastructure, the company built EnVision: a high-performance DCIM (Data Center Infrastructure Management) platform.
The EnVision platform collects real-time electrical and environmental data from equipment worldwide, structures it in a centralized database, and presents it through mission-critical dashboards and reports. Digital Realty’s own operators rely on it daily for global operations, while enterprise customers access their dedicated data segments through tenant portals.
Because this telemetry directly feeds the company’s 99.99% uptime commitments, tenant billing, and capacity planning, SQA was engaged to verify that the platform reports exactly what is happening on the data center floor—securely, accurately, and at scale.
The Challenges
Testing an enterprise-grade platform of this magnitude presented several highly specialized technical and operational challenges:
- Critical Infrastructure Monitoring: The platform monitors the equipment that keeps a data center alive, including UPS units, PDUs, automatic transfer switches (ATS), generators, CRAC/CRAH units, chillers, and HVAC systems. A single wrong reading can mask a developing failure or trigger a costly false alarm, expending operator attention and resources.
- Dual Sources of Truth: Facility data flows in parallel through the Schneider Electric building management system (BMS) and the DCIM platform. Both systems must maintain absolute alignment on power, energy consumption, and environmental parameters, despite collecting data over different paths, using different protocols, and aggregating it differently.
- Multi-Protocol, Multi-Vendor Environment: Telemetry arrives from hardware built by various vendors over several industry protocols, including Modbus, BACnet, SNMP, and branch circuit monitoring (BCM). Each protocol carries its own data formats and unique failure modes, meaning validation had to cover the entire chain from physical device to dashboard.
- Deep Reporting Hierarchy: Operational readings roll up from cabinet to room, floor, site, region, and global levels. This data feeds more than twenty standard reports and four primary management dashboards. Any data corruption or calculation error at the lowest level travels upward into the numbers that critical strategic and operational decisions rest on.
- Global Scale and Multi-Tenancy: EnVision serves both internal operators and external enterprise customers. It ingests high-frequency readings from thousands of sensors across multiple regions while needing to keep its web interface responsive under heavy concurrent user loads. Furthermore, strict multi-tenant isolation was mandatory to prevent any data leakage between customers.
The solution
To address these challenges, SQA designed and implemented a comprehensive testing and validation framework that blended deep functional verification with custom automated monitoring layers:
- Dashboard and Report Testing: We verified the monitoring dashboards and standard reports at every hierarchical level, ensuring that displayed values, historical trends, alarm states, and roll-up calculations perfectly matched the underlying operational data. This included validating report exports across all supported formats (XML, CSV, PDF, Excel, TIFF, Word) and testing the platform’s “Change 1 – Change All” inheritance behavior.
- Metric Validation & Data Reconciliation: Our team validated the raw readings, derived metrics, and threshold alarms for all critical power and cooling equipment against expected hardware behavior. We also established a cross-system data reconciliation pipeline between the Schneider BMS and the DCIM platform, tracing any detected discrepancies back to device configuration, protocol mapping, or aggregation logic.
- Alarm, Notification, and Security Auditing: We verified that critical threshold breaches triggered real-time alarms, email alerts, and SNMP notifications, testing the escalation logic to guarantee prompt delivery. Additionally, we conducted rigorous access rights and data segregation testing across customer portals to ensure flawless multi-tenant isolation.
- Performance and Resilience Testing: We subjected the web interface to load tests with growing numbers of concurrent users and evaluated database performance under high transaction volumes. Failover and disaster recovery tests were executed to validate network load balancing, clustered server configurations, and daily backup procedures under stress conditions exceeding peak operational limits.
- Custom Automation Layer: Recognizing that manual spot checks cannot keep pace with thousands of data points refreshing every minute, SQA engineered an automated validation suite that watches the platform itself. We deployed custom Bash and Python scripts via cron on the Linux servers to poll devices directly using Net-SNMP and Modbus utilities, comparing those values against the SQL databases to flag mismatches. Selenium suites were built to open dashboards on a schedule and verify the UI data against the backend, while Postman collections ran automated checks against the platform’s REST API. When a failure occurred, the scripts automatically harvested relevant syslogs via SSH using grep and awk, allowing the engineering team to begin investigations instantly based on hard log evidence.
The results
The collaboration between Digital Realty and SQA yielded a highly stable, fully verified monitoring ecosystem with concrete operational outcomes:
- Total Data Consistency: Guaranteed reliable, verified monitoring data across both the BMS and DCIM platforms globally.
- Proactive Defect Resolution: Telemetry and data quality issues are now caught and traced to their root cause during development, long before reaching operational decisions or customer billing reports.
- Bulletproof Multi-Tenancy: Confirmed absolute data isolation across customer portals, ensuring enterprise clients can only see their own infrastructure.
- Proven Infrastructure Resilience: Load, failover, and database stress testing proved the platform maintains responsiveness and integrity during peak traffic and hardware failures.
- Continuous Integrity Monitoring: The deployment of repeatable, scheduled automated checks ensures that data drift between hardware devices, databases, and dashboards is caught and flagged 24/7.
