NetScalers decommissioning challenges
Unprecedented scale and complexity
The customer had the third-largest NetScaler footprint globally, but the most configuration complexity of any deployment worldwide. Multiple business units maintained distinct environments, change processes, and dependencies.
High-volume, high-stakes traffic
Each migration affected terabits of live production traffic. Even a minor misconfiguration could cause visible impact across global services and impact hundreds of millions of users.
Tight global deadline
The program required end-to-end migration, from initial audit to hardware decommissioning, within just nine months, across distributed teams and data centers in the U.S. and China.
NetScaler migration process
OpsWerks embedded engineers directly with internal teams to lead a phased, automation-driven NetScaler migration with zero unplanned downtime.
01
Phase 1: Audit and Classification
Audited ~9,300 VIPs across U.S. and China environments (from dev to production) across multiple business units.
Analyzed 60–90 days of live traffic to identify active vs. inactive VIPs for migration or decommissioning.
Developed automation scripts for VIP/device inventory auditing, bulk connectivity testing, and burn-down reporting, saving hundreds of hours and reducing manual errors.
Coordinated ownership validation across dozens of internal service teams.
02
Phase 2: Configuration and Cutover
Built equivalent configurations on the new load balancer and validated parity in QA, performance, and pre-prod before production rollout.
Executed controlled traffic cutovers during maintenance windows with continuous monitoring.
Introduced a “logical decommissioning” process disabling interfaces before final shutdown to enable instant rollback if needed.
Extended OpsWerks’ scope to full production NetScaler migration after proving stability in early phases.
Created dashboards for live traffic visibility, enabling teams to track migration progress and system behavior.
02
Phase 3: Decommissioning and Standardization
Monitored live traffic post-cutover to confirm stable operations.
With local data center teams coordinated safe decommission of 1,600+ physical NetScaler appliances after validation.
Updated and standardized technical runbooks, tribal knowledge, and monitoring documentation for repeatable operations.
