How Vix Technology Solaris Servers Migrated to AWS Using Stromasys Charon SSP Emulator
By Paul Grobler and Gary Googins – Vix Technology
By Sandy Levitt and John Gutknecht – Stromasys
By Rohit Darji and Kevin Yung – AWS
Vix Technology is a global leader in automatic fare collection, transit information, and transit analytics solutions. It has been driving changes in fare collection for over 35 years by building on long-term partnerships with transit agencies and operators.
Vix’s products, people, processes, and pedigree enable it to solve problems and deliver solutions that are changing the world of travel.
For more than 20 years, Stromasys has provided the world’s leading organizations with a better way to break free from the growing risks associated with legacy hardware maintenance. Its Charon emulation software and collaboration with Amazon Web Services (AWS) enable customers to seamlessly virtualize legacy systems to the cloud and eliminate the cost and time needed to manage hardware and rewrite software.
With Stromasys, no re-engineering is required because Charon’s unique capabilities allow it to run legacy applications unchanged on its native operating system in the cloud.
In this post, you will learn the migration best practices and solutions successfully applied in Vix Technology’s Seattle Data Center AWS migration in 2020 and 2021.
Vix’s key business drivers for migration were to move to a more secure, cost effective, and resilient environment, thus reducing outages and incidents for the third-party hosted system.
In this migration, Vix partnered with AWS ProServe and Stromasys, an AWS Migration and Modernization Competency Partner, to successfully migrating over 100 applications, including both x86 and SUN SPARC Solaris mission-critical workloads within nine months before a data center contract renewal with near-zero customer impact.
In 2020, Vix Technology decided to collaborate with AWS ProServe and Stromasys to execute on its ORCA regional fare collection system data center exit plan.
The nine-month migration plan was ambitious, involving two data centers, more than 100 applications, and a complex network of seven independent agencies within the transport system.
The most critical decision was the migration of Vix’s in-house developed accounting application and databases which resided on SUN SPARC Solaris hardware. The complex accounting Clearing House application applies local government pricing rules to hundreds of thousands of daily transactions to clear, settle, and apportion revenue to the agencies on a daily basis.
In making the migration decision, Vix was not sure if the critical application could be migrated to AWS using Stromasys emulation and continue to meet existing performance requirements. In addition, to support Vix’s development in the cloud for its next-generation ticketing system, executives and senior leaders were looking for a good return on investment (ROI) for this migration.
During the critical migration period, the pandemic hit the project team and all travel was suspended. Without on-site collaborations, Vix engineers, AWS ProServe, and Stromasys worked remotely in a joint effort, planning and executing the migration to race through within the nine months migration deadline.
Key to delivery of the migration effort to exit the data center was close collaboration between all parties from the start. The team accomplished the accelerated migration by leveraging the AWS migration framework, namely through the assess, mobilize, and scale migration phases.
In early 2020, before the migration took place, Vix and AWS ProServe started a detailed assessment for the data center migration. The migration assessment focused on providing Vix a directional business case, application dependency analysis, and a migration plan including “7R” decisions. This included migration networking solutions, security recommendations, and recommendation to migrate the Clearing House application using Stromasys Charon SSP emulation.
A proof of concept (PoC) was conducted by the Vix team and Stromasys through an AWS Marketplace deployment, proving the viability of this for a production solution. The AWS Marketplace hourly consumption model and its point-and-click deployment simplicity provided Vix the means to conduct a rapid PoC, allowing the migration decision process to be accelerated and taken within two weeks.
Meticulous planning and execution by the team ensured the ORCA migration to be a smooth transition with minimal customer impact. The accuracy of the discovery phase set a strong foundation for the project, and the AWS team ensured a timely, controlled delivery to exit the data center ahead of contract renewal.
During the discovery, Vix engineers proved the Clearing House application could be lifted and shifted into the Stromasys emulator hosted on an Amazon Elastic Compute Cloud (Amazon EC2) instance. This was achieved by taking a snapshot of the SUN SPARC server in the test environment and restoring it into the Stromasys emulator.
Furthermore, by running thorough stress testing, the team verified the migrated application performance was able to complete end-of-day (EOD) processing in under 90 minutes, a requirement defined by existing hardware (Sunfire V490, 8 CPUs, and 16GB of memory). In the AWS PoC setup, the Vix engineer deployed the Clearing House application on to a z1d.6xlarge EC2 instance installed with the Stromasys Charon SSP emulator.
AWS Solution Architecture
The AWS architecture for the Clearing House application is illustrated below.
Figure 1 – Target solution architecture.
In designing the architecture for the migration, the team considered the following best practices and trade-offs:
- The application needed to mimic the active-standby setup of the on-premises setup as it operated across the two data centers. As a result, the design used multi-AWS Availability Zone deployment with two z1d.6xlarge instances.
- It’s best practice to reduce changes and test complexity during a short-term migration. As a result, the migration took the existing file system backup of the SUN Solaris servers and restored it in the Stromasys emulator using virtual disks. This approach allowed the team to predominantly reuse existing installed dependencies and operational scripts.
- To meet the 24-hour recovery time objective (RTO) and recovery point objective (RPO), the design adopted an automated nightly backup with Amazon Elastic Block Store (Amazon EBS) volumes. During a failover event, a Vix engineer can restore the nightly snapshot to a standby instance and perform an application failover.
- To meet the performance requirement for the Clearing House application database, the gp3 EBS type is used with provisioned IOPS.
Learnings and Best Practices
Compounding the short timelines and accelerated approach required to hit the nine-month data center exit deadline, the pandemic further added to the challenge.
The team overcame technical challenges, but more importantly demonstrated that customers can achieve great migration velocity and success with support in technologies and specialists from the AWS Partner Network (APN) and AWS ProServe.
To summarize, the following are key learnings from the migration:
- In the early stage of the engagement, AWS ProServe conducted a migration assessment and discovery. By replaying the assessment results, Vix was able to understand the migration business value and gain better visibility in the migration patterns and plan. During the discovery, the team mitigated a migration risk by conducting a rapid PoC to prove the technology.
- Vix leveraged Stromasys’s AWS Marketplace offering to do a low-cost PoC by using a pay-by-hour subscription fee. Vix continued to use the Marketplace AMI for its production deployment, reusing the blueprint from the PoC, and this reusability accelerated the migration. Stromasys also provides virtual environment (VE) license deployment option for customers who prefer to use more cost-effective annual licenses.
- Best practice was followed by performing the short-term migration with minimal application changes during the migration phase to achieve cost savings, which in turn could be used for long-term modernization and innovations.
- Vix optimized AWS operating cost savings by subscribing to three years of reserved instances (RIs) for the migrated workloads.
- Use of tools-based migration facilitated quick and error free migration.
Additionally, from an organizational perspective, Vix noted that sufficient capacity internally needed to be available to support the full application migration effort. This needed to be planned into the migration. Vix also noted that internal teams should be trained and prepared to understand the benefits of the migration early in the project.
Vix Technology’s successful exit of its data center shows legacy SUN Solaris servers are not a blocker for migration to AWS. By working as one team together with Stromasys and AWS, Vix exited the data center on time, achieving an overall saving of 40% versus on-premises cost. In addition, Vix achieved its further goals of faster process times within a more secure, resilient environment.
If you want to migrate applications to AWS but aren’t sure how to migrate the mission-critical applications running on SUN SPARC or PA-RISC servers, reach out to Stromasys at email@example.com, or visit the Charon on AWS website.
If you’re looking for solutions to migrate critical workloads in mainframe and midrange to AWS, contact AWS Mainframe Modernization Business Development at firstname.lastname@example.org.
Special Thanks: This migration success is a result of close collaboration between Vix Technology, AWS, and Stromasys. We want to thank Vix executives’ strong sponsorship, and the dedication from AWS account team members Lynn Macmillan and Juliana Mahecha; AWS ProServe team members Simone Kobelke, Prashit Dhingra, and Ward Britton; and Stromasys team members Thomas Netherland, Phil Smith, Alexandre Souf, and Kevin Kelly.
Stromasys – AWS Partner Spotlight
Stromasys is an AWS Migration and Modernization Competency Partner that provides cross-platform virtualization (emulation) software for legacy systems.