High Availability: Keeping EUDI Wallet Services Running 24/7

High availability (HA) is a characteristic of a system that aims to ensure an agreed-upon level of operational performance -- usually expressed as uptime percentage -- for a higher-than-normal period. Systems with 99.99% availability ("four nines") experience no more than 52.6 minutes of downtime per year. For the EUDI Wallet ecosystem, high availability is a non-negotiable requirement because digital identity infrastructure supports critical services including border control, healthcare access, financial transactions, and government interactions for hundreds of millions of EU citizens. An outage of credential verification services could have cascading effects across multiple sectors and Member States simultaneously.

Availability is measured as the percentage of time a system is operational. Small differences in percentage translate to significant differences in allowed downtime:

• •99% (two nines): Up to 3.65 days of downtime per year. Insufficient for critical identity services.
• •99.9% (three nines): Up to 8.76 hours per year. Acceptable for non-critical supporting services.
• •99.99% (four nines): Up to 52.6 minutes per year. Target for core EUDI verification services.
• •99.999% (five nines): Up to 5.26 minutes per year. Aspirational target requiring extreme redundancy.

Each additional nine of availability requires significantly more investment in redundancy, monitoring, and operational procedures. EUDI service operators must balance availability targets against cost and complexity, guided by the criticality of each service component to the overall identity ecosystem.

EUDI Wallet infrastructure employs several well-established HA patterns:

Active-active clustering: Multiple instances of each service run simultaneously across different servers or regions. Traffic is distributed among all instances. If one fails, the remaining instances absorb its load. This is used for stateless services like API gateways and credential verification endpoints.

Primary-replica database architecture: A primary database handles writes while multiple replicas handle reads. If the primary fails, a replica is automatically promoted. This provides both HA and read scalability for data-intensive operations like Trusted List queries.

Multi-region deployment: Services are deployed across at least two geographically separated EU data centers. Global load balancing directs users to the nearest healthy region. This protects against regional failures (power outages, natural disasters) and provides lower latency for users across different parts of Europe. Data sovereignty requirements are respected by keeping all data within EU borders.

Maintaining high availability requires complete monitoring and defined Service Level Agreements (SLAs). EUDI service operators implement:

Real-time dashboards showing current availability status for each service component, request latency percentiles, error rates, and infrastructure health. These dashboards are monitored 24/7 by operations teams.

Alerting systems that notify teams immediately when availability drops below threshold levels, latency exceeds acceptable bounds, or error rates spike. Alerts are tiered by severity, with critical alerts triggering immediate response procedures.

SLA reporting provides monthly and quarterly availability reports to regulators and stakeholders. The eIDAS 2.0 framework requires EUDI service operators to maintain defined availability levels and demonstrate compliance through regular reporting. Failure to meet SLA targets may result in corrective action requirements or, in severe cases, certification review.