Skip to main content

Introduction

PolyScale.ai - The Global Database Edge Cache​

PolyScale makes databases fast, everywhere.

PolyScale is a Platform as a Service that makes data-driven apps faster by simplifying global data distribution and caching. PolyScale caches database data locally to users reducing latency, accelerating read performance and scaling throughput.

Using PolyScale's global network (Points of Presence), applications use local database query-compute and storage, reducing latency, accelerating read performance and scaling throughput.

What Is PolyScale?​

At its core, PolyScale is a global database cache as a service. Using PolyScale, database data can be distributed and cached, seamlessly scaling your current database without altering transactional semantics.

PolyScale's global network proxies and caches native database wire protocols. It transparently connects to your current database and requires no code or infrastructure changes. Queries are inspected and reads (SQL SELECT) can be cached geographically close to the requesting origin for accelerated performance. All other traffic (INSERT, UPDATE and DELETE) seamlessly pass through to the source database.

PolyScale provides a plug and play approach to global data distribution and scaling.

The Data Distribution and Scaling Challenge​

With the adoption of globally replicated delivery (CDN) and distributed computing frameworks (Edge serverless), global latencies continue to be reduced. Users expect low latency performance no matter where they are or how complex the underlying requirements may be.

Today, it is relatively simple and inexpensive to deploy both static assets and business logic locally to users, pushing latencies lower. This paradigm shift however poses challenges at the data tier in that storing consistent data in a distributed manner implies additional operational overhead, cost, and strains monolithic databases.

Data latencies are typically reduced for users by scaling the database (for example adding read-replicas), or by implementing application level caching using in-memory solutions such as Redis. Both of these solutions come with ongoing costs. Distributed databases can help, but the distributed aspect makes them complex to deploy and maintain.

Benefits of PolyScale​

The addition of a PolyScale cache provides several benefits:

  • Increased Query Performance - PolyScale serves any cached queries sub-millisecond with massive concurrency on multi-terabyte data sets. Any slow queries are served instantly without the load burden on the database.
  • Lowered global latency - PolyScale provides a global edge network so that data can be cached locally to wherever your application is running. Connecting to PolyScale automatically resolves to the closest Point of Presence, so data is fast everywhere.
  • Reduced database infrastructure costs - read queries are served from Polyscale rather than the database. PolyScale is a serverless platform and hence you only pay for what you need, reducing the need for infrastructure that has been sized for peak traffic workloads. Read replicas are also often removed from architectures in favor of Polyscale.
  • Increase engineering productivity - Unlike caching databases such as Redis, PolyScale requires no code development or configuration. This removes the requirement for cache design entirely. No data modeling, invalidation design, testing and observability. Developers can focus on application features rather than scaling their database.
  • High availability and fault tolerance - PolyScale's global edge network provides inbuilt fault tolerance so that in the unlikely event of an network outage, the next closest Point of Presence is automatically used.