The Heimdall Proxy is a data access layer for application developers, database administrators, and architects. Whether on-premise or cloud, our proxy deliver a faster, more scalable, and secure solution for their current SQL database.
How is the Heimdall Proxy Unique?
The Heimdall Proxy gives users visibility and control over their database. The proxy can be is deployed as a transparent sidecar process. Our distributed model results in optimal performance and predictive scale.
Automated Query Caching: Today’s SQL solutions incur data latency from application-database round trips. Caching on the database improves scale, but does not remove the latency. With the Heimdall Proxy caching, queries that would have gone to the database are now cached before hitting the database. You choose the cache of choice (e.g. Amazon Elasticache, Redis, Hazelcast, GemFire) and we determine which query results to cache and invalidate. Best of all, deployment does not require writing a single line of code.
Read/Write Split with ACID Compliance: To horizontally scale the database tier with master writer and read replicas, application modifications are required. Our proxy is routes queries to the appropriate database instances with replication lag detection to ensure data consistency.
Advanced Connection Pooling: Connections to the database are reduced by up to a 1000:1 ratio. Connections are controlled per user and per database, ensuring fairness while protecting the database from being overwhelmed (respectively).
Comments on Heimdall Proxy’s uniqueness:
- Granular 1) Per user and 2) Per database server connection limits
- Read/write split with replication lag detection to ensure fresh data
- Database vendor-neutral
- Pg-Pool-II and ProxySQL claim support for Read/Write Split and Cache Invalidation. However, they do not guarantee data consistency, and cannot be supported on multiple proxies as a cohesive unit. Check out our webinar.