Upstash as a Serverless Database for Edge

Upstash embarked on its mission to be an optimal database for AWS Lambda functions, but it identified an opportunity to serve Cloudflare Workers, which offer global low-latency execution but lack stateful data management. To solve this, Upstash developed solutions like a REST API tailored for Cloudflare Workers and edge caching, which significantly improved performance by caching Redis® REST responses at edge locations. However, edge caching presented limitations like stale data and support only for REST API, prompting Upstash to introduce a Global Database solution that replicates data across multiple AWS regions, offering eventual consistency without cache-invalidation issues and making it a suitable option for both read and write operations. Upstash's Global Database is designed to minimize read latencies while edge caching is recommended for cases where caching is sufficient, with the possibility of combining both solutions for optimal performance. Looking ahead, Upstash plans to enhance its architecture for low-latency writes and expand its capabilities by introducing Kafka support to facilitate use cases such as clickstream analysis and serverless function logging.