DoorDash Scales Global Reliability with Cortex XCOR Cloud-Native Observability

SUMMARY

DoorDash is the global leader in on-demand delivery, connecting hundreds of millions of consumers with local merchants worldwide. Its software platform drives the core customer experience so any system outage directly impacts order fulfillment, brand reputation and global revenue.

To protect uptime — even during high-stakes demand spikes like the Super Bowl and Valentine's Day — DoorDash partnered with Palo Alto Networks to use Cortex^® XCOR^™. The solution supports DoorDash’s SLO-driven observability practice, aligning software health and reliability with business outcomes. Cortex XCOR also optimized 92% of telemetry by pruning low-value data, lowering total cost of ownership (TCO) and freeing engineering teams to focus on high-signal data.

RESULTS

70,000

automated SLOs scaled reliability across all critical engineering workloads.

92%

data optimization reduced observability spend while preserving key telemetry.

325,000

recording rules powered real-time operational visibility across global traffic.

THE GOAL

Scaling Observability Across Massive Traffic Spikes

DoorDash connects hundreds of millions of consumers with local merchants worldwide, where software reliability directly impacts merchant revenue and consumer trust. High-demand events like Valentine’s Day and the Super Bowl underscored the need for observability infrastructure that could scale efficiently with DoorDash’s traffic and telemetry volumes.

Additionally, as the business grew and DoorDash expanded its microservices architecture, managing high cardinality and escalating telemetry costs across thousands of services became a key engineering priority. DoorDash aimed to make reliability goals consistently measurable across teams, giving business and technical leaders a shared view of service health and overall system reliability.

SOLUTION

Automated SLO Governance and Control Plane Efficiency

DoorDash equipped engineers with turnkey SLO automation built directly into service code for ease of use, establishing a tiered reliability framework across all applications. These 70,000 SLOs carry over to the Cortex XCOR platform for monitoring service health, creating a shared operational language that connects technical metrics to business health and global order fulfillment.

To control escalating observability costs, DoorDash utilized the Cortex XCOR Optimization Engine. Based on the solution’s automated recommendations, DoorDash’s Observability Team implemented over 325,000 recording rules to support service monitoring and used aggregation rules to achieve a 92% data optimization rate, reducing low-signal noise and telemetry overhead. Along with lowering costs, this approach equips the engineering team with critical real-time visibility to keep DoorDash services available and reliable.