Investigate caching strategies

[PatStLouis]

Explore Caching Strategies for Performance Optimization

Context

The DID WebVH Server handles high-frequency DID resolution requests and database queries. Implementing caching will improve response times and reduce database load.

Current Pain Points

• DID resolution endpoints are frequently accessed
• Database queries for policies and registries on every request
• No caching layer currently implemented
• Potential performance bottlenecks as traffic scales

Caching Options to Explore

Option 1: Redis Cache (Distributed)

Best for: Multi-instance deployments, horizontal scaling

• ✅ Shared cache across instances
• ✅ Fast in-memory performance
• ✅ Supports TTL and complex invalidation
• ❌ Requires Redis infrastructure
• ❌ Additional operational overhead

Use Cases:

• DID document resolution
• Policy/registry lookups
• Cross-instance cache sharing

---

Option 2: In-Memory Cache

Best for: Single-instance deployments, development

• ✅ No external dependencies
• ✅ Fastest access (no network)
• ✅ Simple implementation
• ❌ Not shared across instances
• ❌ Limited by server memory

Use Cases:

• Development/testing
• Single-instance deployments
• Hot data caching

---

Option 3: HTTP Response Caching

Best for: Public endpoints, static content

• ✅ Reduces server load entirely
• ✅ Leverages browser/CDN caches
• ❌ Only for GET requests
• ❌ Less control over invalidation

Use Cases:

• Public DID resolution
• Explorer UI assets
• Static content

---

Cache Invalidation Strategies

• TTL-based: Automatic expiration (simplest)
• Write-through: Update cache on writes (consistent)
• Pattern invalidation: Delete related keys on updates
• Version-based keys: Natural invalidation via version numbers