AWS re:Invent 2025 - Ticketmaster: Enhancing live event experiences for fans with AWS (SPF206)

Enhancing Live Event Experiences with AWS Local Zones: Ticketmaster's Journey

Ticketmaster's Background and Journey with AWS

• Ticketmaster is a 50-year-old live event ticketing and management company
• Sells over 600 million tickets annually across 35+ countries
• Handles ticket sales, event management, and entry validation for customers
• Began adopting AWS about 10 years ago, starting with basic cloud services
• Evolved to become a heavily hybrid organization, with workloads split between on-premises and AWS

The Importance of Latency for Ticketmaster's User Experience

• Ticketmaster's website and applications require sub-millisecond latency for features like:
  • Interactive seat maps and seat selection
  • Real-time ticket inventory and recommendations
  • Seamless integration of on-premises and cloud-based systems
• Latency issues can significantly degrade the user experience, causing frustration and abandoned transactions

Ticketmaster's Networking and Infrastructure Evolution

1. Pre-Cloud: Ticketmaster had data centers on the East and West coasts of the US, with applications and databases co-located. This worked well for low-latency, but had limitations for global reach.
2. Early Cloud: Ticketmaster chose the AWS US East region, which provided low latency to their East Coast data center. For West Coast workloads, they had to choose a region further away, resulting in 30-40ms latency.
3. Local Zones: About a year ago, Ticketmaster started working with AWS on Local Zones, which provided sub-2ms latency connections between their Phoenix data center and the local AWS infrastructure.

Architecting for Low Latency with AWS Local Zones

• Ticketmaster's low-latency architecture leverages:
  • Direct Connect from on-premises to Local Zones
  • Transit Gateway to connect Local Zones to the parent AWS region
  • VGW (Virtual Gateway) to enable low-latency east-west connectivity between Local Zones
  • Separate VPCs for each Local Zone to optimize routing
• This allows Ticketmaster to:
  • Burst capacity into the cloud during high-demand events
  • Flexibly place workloads in the optimal location (on-prem, Local Zone, or parent region)
  • Achieve sub-millisecond latency for critical user-facing features

Best Practices and Lessons Learned

1. Treat Local Zones as you would Availability Zones in a region.
2. Carefully evaluate the service and instance type availability in each Local Zone.
3. Ensure you have a dedicated Direct Connect from on-premises to the Local Zone, not routed through the parent region.
4. Deploy your own firewall appliances in the Local Zone, as managed services may not be available.
5. Use separate VPCs for each Local Zone to optimize routing and connectivity.

The Road Ahead for Ticketmaster

• Expanding the use of EKS Hybrid Nodes to enable greater workload placement flexibility
• Exploring the use of Local Zones in other regions to address data residency requirements and bring the cloud closer to end-users
• Continuously evaluating new AWS services and capabilities in Local Zones to enhance Ticketmaster's live event experiences