Improving Cost Efficiency by 40% Using Amazon Managed Service for Prometheus with Choice Hotels

Choice Hotels is a large lodging franchisor that has more than 7,400 hotels, representing over 625,000 rooms, in 45 countries and territories. Choice Hotels has consistently been at the cutting edge of new technologies, providing its customers and hotel owners with support and innovations. Aiming to migrate entirely to the cloud by 2024, Choice Hotels was one of the first major hotel companies to commit to AWS.

Choice Hotels uses Prometheus as its metric observability solution, recording about 25 million metric series per minute. The company also used several other solutions and still faced challenges. First, the solutions—especially the proprietary tracing one—needed to support the full workload. Second, Choice Hotels needed standardization between its observability systems so that no service would be named X in one system and Y in another. Choice Hotels recognized these issues and sought a simpler, standardized approach that could support its workload and drive interoperability between technologies.

Originally, the team had planned to fully migrate the rate, error, and duration metrics pulled from more than 15 million spans per minute to its self-managed Prometheus. However, with the high volume of metrics involved, Choice Hotels would need to massively scale and spend much time managing the data. In 2020, Choice Hotels migrated to Amazon OpenSearch Service—which provides near real-time search, monitoring, and analysis of business and operational data—and saw improvements in capacity, cost, and performance. To replicate these benefits with another managed solution, Choice Hotels decided to use Amazon Managed Service for Prometheus, a service that monitors and provides alerts on containerized applications and infrastructure at scale.

“We subscribe to the same principle that AWS sticks to, which is to get rid of undifferentiated heavy lifting,” says Jason Simpson, vice president of engineering at Choice Hotels. “Using the managed service, we can achieve that.”

Choice Hotels began migrating its tracing observability solution in May 2022. In addition to Amazon Managed Service for Prometheus, it began using AWS Distro for OpenTelemetry (ADOT), which provides open-source APIs, libraries, and agents to collect distributed traces and metrics for application monitoring (see figure 1). Rolling out the right solution and proper configuration took time and effort, but working alongside the AWS team, Choice Hotels’ cloud team quickly resolved issues. “AWS gave us architectural support for the design to verify that we weren’t doing anything against best practice,” says Stephen Cihak, senior director for cloud platform at Choice Hotels.

Using Amazon Managed Service for Prometheus and ADOT, Choice Hotels met its goal of standardization. With a standardized dashboard for its rate, error, and duration metrics, Choice Hotels now has a more simplified view into its system. The system is interoperable with the tracing visualization, and the troubleshooting process is smoother. “We know where to go; we know what to look for,” says Stephen. “I can now pull up related traces with the click of a button.”

Migrating to a managed system reduced the maintenance time. The system has optimized the team’s performance, freeing up time and other resources. “Upgrading and managing self-managed systems can be complicated and error prone,” says Stephen. “Now, with a managed service, we only go in and press a button, and we don’t have to worry about it.” In addition, Choice Hotels improved cost efficiency by 40 percent after migrating away from proprietary and self-managed solutions and optimizing how the team batches metrics.

On AWS, Choice Hotels also gained the agility to continue improving performance and optimizing cost. Before, Choice Hotels would have had to overhaul its system to improve capabilities or cost. Now, with a managed system, it can implement new technology or adopt mechanisms without going through a migration process or building a high level of expertise. “This is definitely a good long-term plan for us,” says Jason. “One of our overall strategies is to adopt new tech as it comes out quickly. The value of AWS is that we can make continual improvements.”