AWS Database Blog
How Omnissa saved millions by migrating to Amazon RDS and Amazon EC2
Omnissa, formerly the VMware End-User Computing business, is a digital workspace technology leader that delivers smart, seamless, and secure digital work experiences for organizations worldwide. It serves 26,000 customers, including the top seven of the Fortune 500 companies. The company’s mission is to simplify and secure the digital work environment, empowering employees to work productively from anywhere, on any device, without sacrificing IT control or user experience. Omnissa Workspace ONE®, Unified Endpoint Management (UEM) , helps its customers secure and manage all devices that connect to a network, both on premises and in the cloud. Omnissa Horizon® provides solutions for virtual desktops and apps, digital employee experience, and robust security—all underpinned by AI-driven automation.
With AWS, Omnissa was able to successfully migrate its mission-critical workloads from VMware Cloud on AWS (VMC-A) to native AWS services, including Amazon Elastic Compute Cloud (Amazon EC2) and Amazon Relational Database Service (Amazon RDS) for SQL Server. This migration exemplifies the strategic impact of Omnissa’s cloud transformation journey. Using these services, it not only achieved a substantial cost reduction of 39%, saving millions, but also significantly enhanced operational efficiency.
Omnissa used this migration as an opportunity to start a transformation initiative, which included refactoring its applications and databases as part of a multi-year modernization strategy. This approach aims to further optimize costs and transfer the resulting savings to its customers.
Josh Broch, VP of Engineering at Omnissa, commented on the success of the migration:
“The transition from VMC-A to Amazon RDS for SQL Server and Amazon EC2 was a critical strategic move for us. The AWS team’s expertise, proactive approach, and unwavering support were instrumental in making this complex migration a success. Not only have we achieved significant cost savings, but we’ve also gained a more flexible, scalable infrastructure that allows us to better serve our customers and drive innovation.”
The Omnissa divestiture and challenge
Omnissa became an independent business in July 2024, when the VMware End User Computing (EUC) business was divested to KKR, a global investment firm. VMware had previously been acquired by Broadcom and announced potential price increases for VMware Cloud on AWS services. This put Omnissa in a challenging position. Its flagship product, UEM platform ran on VMC-A across ten AWS Regions, representing a significant operational footprint and financial commitment. The Omnissa UEM Engineering team needed to find a cost-effective solution that wouldn’t compromise the quality and performance of its services so it could continue to provide the best experience for its global customer base.
After careful consideration, Omnissa determined that migrating its self-managed SQL Server databases to Amazon RDS for SQL Server and application servers to Amazon EC2 would solve this challenge. However, before the team could begin migrating its workloads, it needed to build a business case with leadership.
Building a business case for AWS migration
To start building a business case, Omnissa’s UEM Engineering team worked with the AWS team to begin assessing the current workloads. As an initial step, Omnissa used RVtools, a lightweight third-party application to export the current VMC-A environment data. Once it had been exported, AWS used this information to assess Omnissa’s workload and licensing requirements. They segregated self-managed SQL databases environments into those requiring SQL Server Enterprise Edition-specific features and those needing only SQL Server Standard Edition. This targeted approach ensured that each database environment used the edition best aligned to its requirements.
The AWS team conducted in-depth analysis of the current and historical usage patterns for the Omnissa application and database instances. The analysis revealed the company’s database workload was characterized by memory-intensive operations with high temporary storage needs, but relatively low CPU utilization on its current instances. It also revealed that the application workload could be less compute intensive. This triggered the hunt for the right choice of instance to optimize costs.
This analysis confirmed that using Enterprise Edition on RDS for SQL Server (AWS provided license) and self-managed SQL Server on EC2 (customer provided license) aligned with Omnissa’s requirements. The Omnissa UEM Engineering team opted to defer the adoption of Amazon Aurora PostgreSQL-Compatible Edition as part of the future modernization roadmap. Omnissa’s urgent need to complete this migration before the impending VMware licensing cost spikes on its existing infrastructure drove this decision.
The AWS team helped develop a business case comparing the two options and identified significant optimization opportunities that the Omnissa team took advantage of to reduce licensing costs by following the recommendations below:
Strategic instance selection: For the database workload using memory-optimized instances, like x2iedn instances, Omnissa minimized instance costs by reducing the number of cores used without exceeding the original memory requirements. For application hosting, the optimal choice was the compute-optimized c6i instance class, offering enhanced performance, improved flexibility, and broad Regional availability. Choosing the right instance type and size helped to reduce Omnissa licensing expenses by 66%.
Optimizing SQL Server editions: By upgrading database workloads to the latest SQL Server editions, Omnissa could take advantage of the latest features. Using RDS for SQL Server, it could downgrade databases from Enterprise to Standard Edition and utilize built-in RDS features to replace Enterprise-only capabilities while reducing costs. However, Omnissa preferred to use SQL Server Enterprise Edition for feature usage flexibility and simplified operations.
Automate database administrative tasks: With RDS for SQL Server, Omnissa could automate time-consuming database administrative tasks, such as push-button scaling, automated provisioning, automated patching, and built-in high availability and disaster recovery (HADR) capabilities. Omnissa would also be able to natively integrate with AWS AI/ML services.
Reducing total cost of ownership: By offloading these operational activities, Omnissa DBAs could focus on collaborating with architects and developers on strategic initiatives, like modernization.
Migration & Modernization programs. Through the AWS VMware Migration Accelerator (VMA) program, Omnissa would receive migration credits that offset parallel environment costs during its transition from VMware Cloud to EC2 and RDS for SQL Server, enabling a cost-effective migration.
This thorough analysis built a compelling business case for the migration, and the assessment revealed potential annual savings of 39% through migration to AWS services. This finding became the cornerstone of the proposal and a key driver for the company’s decision to proceed with the migration.
Planning and capacity management
After the Omnissa UEM Engineering team received approval from leadership, the migration planning stage began. The migration scope included tens of thousands of frontend and backend instances across ten AWS Regions. To streamline operations, Omnissa standardized on a single instance type and SQL Server edition.
Capacity management was a big undertaking given the Omnissa migration included adoption of thousands of newly launched instance types across ten Regions. AWS Enterprise Support proactively ensured the availability of instances weeks before the production migration. To ensure a smooth transition, the customer planned the migration in phases. It started with test and development environments to validate performance and configuration, then moved to production workloads once capacity and scaling requirements were fully assessed.
Migration execution and timeline
The technical implementation of this migration involved several transformative changes, supported by the AWS VMware Migration Accelerator (VMA) program. The program’s tools and methodologies helped reduce migration time and risk while running parallel environments. A significant portion of the project focused on migrating from self-managed SQL Server to RDS for SQL Server using SQL Server’s native backup and restore features. The migration process involved performing full backups followed by differential and log backups on the source instance, then restoring these to the target RDS for SQL Server instance. This approach significantly reduced application cutover time during the migration process. For detailed technical steps on how you can replicate this migration approach, refer to this AWS database prescriptive guidance. This shift was more than a lift-and-shift operation; it represented a fundamental change in how Omnissa managed its databases. The migration to a managed service reduced operational overhead while providing high availability and performance at a lower overall cost. The AWS database specialists worked closely with the Omnissa Engineering team, hosting immersion days and providing hands-on guidance to ensure a smooth transition.
The application stack was migrated using infrastructure automation, and Omnissa Engineering had invested additional time into testing staging and pre-production environments. This approach enabled Omnissa to modernize its infrastructure with minimal operational disruption.
Technical challenges and best practices
As the Omnissa UEM Engineering team began migrating the production workloads at scale, it observed a few challenges such as latency issues and failures during RDS instance up-sizing, user databases becoming stuck in a SUSPECT state, database restore errors, and high memory usage during peak transaction workloads. The application exhibited distinct usage patterns, with peak activity occurring at the start of the workday when users began logging in.
To address these challenges and optimize the configuration, AWS Support and specialist teams collaborated closely with the Omnissa UEM database and migration teams to implement crucial stability improvements. A key optimization involved increasing the Amazon Elastic Block Store (Amazon EBS) gp3 storage volume baseline IOPS from 3,000 to 6,000 for selected instances and raising the SQL Server worker threads beyond the default automatic setting to approximately 1,500 or more, depending on the instance size, to improve parallelism and overall performance. The team established a methodical approach, ensuring IOPS scaling completed before initiating any RDS instance class upgrades. Additionally, Omnissa UEM Engineering team developed a process to identify and manage long-running transactions prior to upgrades, significantly reducing potential disruptions. To maintain visibility into system performance, Omnissa implemented comprehensive monitoring through Amazon Managed Grafana, tracking essential metrics including IOPS, memory, and CPU utilization. The solution was further enhanced by using Query Store on RDS for SQL Server and SQL Server 2022’s advanced observability features, providing deeper insights into database performance.
These combined efforts resulted in a more stable and reliable database infrastructure, particularly during high-demand periods.
Operational improvements
After the migration was complete, the Omnissa Database team looked for opportunities to further optimize performance by running index rebuilds weekly instead of daily, taking new snapshots before upgrades, and exploring the Intelligent Query Processing feature on SQL Server 2022. On application side, the company’s UEM engineering team proactively reviewed the top resource-consuming SQL queries to identify potential application issues and perform technical fixes, such as increasing IOPS, rightsizing, and monitoring to provide quick performance wins. The Omnissa team achieved long-term platform stability through sustained engagement from AWS account team and specialist teams to address operational and scaling challenges and stay on schedule with the migration.
The Omnissa UEM team continues to optimize the migrated workloads by enhancing its runbooks and standard operating procedures. It uses AWS Compute Optimizer for ongoing rightsizing guidance and cost optimization for EC2 and RDS databases.
The regular cadence of technical meetings between Omnissa and the AWS specialist and account teams played a vital role in ensuring a successful migration.
Results and benefits
The results were transformative for Omnissa. Beyond the impressive 39% in annual cost savings, the migration to AWS managed services improved operational efficiency. This delivered savings towards licenses and provided greater flexibility and control over its technology stack. Most importantly, by moving to AWS services, the company’s UEM Engineering team is now well positioned to innovate faster and modernize further. It is planning to enhance the technology ecosystem by migrating most of the databases to Aurora PostgreSQL-Compatible Edition and modernization of its application framework from monolith to micro-services. These strategic moves will significantly reduce licensing costs while establishing a robust platform for sustainable growth and technological advancement.
Conclusion
In this post, we walked through the Omnissa UEM engineering team’s journey of migrating its mission-critical UEM platform and self-managed SQL Server workloads from VMware Cloud on AWS (VMC-A) to Amazon RDS for SQL Server and its application servers to Amazon EC2. For organizations facing similar challenges, whether from rising licensing costs, aging infrastructure, or business transformation events like divestitures, the Omnissa journey offers a valuable blueprint. This migration success demonstrates that with proper assessment, planning, and execution, even complex environments with tens of thousands of servers across multiple Regions can be migrated efficiently and with minimal disruption.
To learn more about how AWS can help your organization achieve similar results, contact your AWS account team or visit the AWS migration resources page.