Blog

Six reasons to cache Snowflake with the Aerospike Database

Matt Bushell
Matt Bushell
Sr. Director, Product Marketing
August 31, 2023|5 min read

Introduction

In the world of modern data management, Snowflake has earned its reputation as a leading cloud-based data warehousing platform, enabling businesses to store, process, and analyze vast amounts of data. Yet, not many are aware of the hidden gem that can supercharge Snowflake’s performance: the Aerospike Database. By integrating Aerospike into their data infrastructure, organizations can reduce latency and increase availability of data needed for real-time decisioning. In this blog post, we’ll delve into the reasons why Aerospike is the perfect companion to Snowflake, unleashing the greater potential of data stored in Snowflake.

1. The Aerospike Database is purpose-built for real-time, mission-critical applications

Snowflake is a data warehouse whose purpose is to serve up data to enable large-scale analytics versus real-time operational data processing – which is the domain of Aerospike. Aerospike is a high-performance, globally distributed, multi-model NoSQL database designed to deliver ultra-fast data access, low latency, and high availability. It is purpose-built for real-time, mission-critical applications including but not limited to customer 360 recommendations and personalization, fraud prevention, AI/ML processing and feature store serving at the edge, real-time risk assessment, mobile gaming, real-time adtech/martech apps (e.g. real-time bidding and ad serving), and internet of things, to name a few.

2. Aerospike “supercharges” caching for reads…and writes

Caching is a fundamental technique used to improve data access speed and reduce the load on the Snowflake data warehouse. By deploying Aerospike alongside Snowflake, businesses can further leverage the data in Snowflake, creating an intelligent caching layer that accelerates operational workloads, providing near-instantaneous access to petabytes of data. However, most caching solutions are typically read-only if not highly read-centric. Aerospike, on the other hand, is able to not only read at sub-millisecond speeds, but can also ingest data at extremely high speeds. By handling mixed workloads, data cached on Aerospike operates modern real-time applications, which in turn leverages the data stored in Snowflake. Aerospike can also enrich cached with its ability to stream data in.

3. Aerospike’s low latency plus high availability is a game-changer

Latency is a crucial factor for operational applications, but they must also be highly available – and this combination is an area where Aerospike excels. Aerospike’s data-centric architecture not only ensures extremely low read and write latencies, typically sub-millisecond, but it is very robust.Aerospike’s high availability stems from its built-in replication and distribution capabilities providing automatic failover and data redundancy. Even in the face of hardware failures or network disruptions, Aerospike’s active-active multi-datacenter deployment options maintain uninterrupted access to cached data.

4. Aerospike’s infrastructure efficiency keeps costs low

Aerospike has a unique, highly-patented hybrid memory architecture that has the performance of in-memory systems with the persistence of flash. Aerospike’s storage engine technology treats flash like memory, without needing to go through the OS enabling highly parallelized access and block reads and writes. Instead of relying on DRAM for performance, Aerospike can rely on Flash, saving 80% on infrastructure footprint and costs.

5. Aerospike scales from gigabytes to petabytes with no performance degradation

As data volumes grow, both Snowflake and Aerospike are equipped to handle increasing workloads. Snowflake’s cloud-native architecture allows for effortless scaling, while Aerospike delivers the ability to readily scale from gigabytes to petabytes with no application changes and no performance degradation. This is due in part to our Smart Clients knowing where every partition of data resides, so every node knows what every other node has. This facilitates a single-hop to the data, regardless of cluster (or dataset) size.

6. Aerospike is proven with Snowflake in real-world use cases

Numerous businesses have successfully leveraged the power of Aerospike to enhance Snowflake data caching. Industries such as e-commerce, financial services, gaming, and ad tech have witnessed substantial improvements in application performance, user engagement, and customer satisfaction.

Case study: Top global asset company caches Snowflake data with Aerospike for advanced analytics

A top global asset company had been using Snowflake heavily for their data warehouse needs. However, they were challenged with ingesting and serving data from Snowflake in real-time (or even near real-time). They were looking to improve their personalization and market risk impact and other advanced analytics applications which had data directly in Snowflake. Their goal was to have real-time decision support based on the sizable amount of data these applications required.

Moving these applications to Aerospike, while complementing Snowflake, solved this challenge for them. The firm created and stored machine learning features and models built on data in Snowflake, conducted inferencing on them, and then batched those features over to Aerospike with data APIs on top of Aerospike to access.

Aerospike was able to pull multiple data sources in real-time including transaction data, market data, and interaction data (chats, calls, web clicks.) In short, Aerospike did not serve as just real-time ingest of data, but a low-latency datastore that sits on top of the Snowflake enterprise data platform for the firm.

Conclusion

As organizations strive to gain a competitive edge in a data-driven world, combining Snowflake with the power of Aerospike Database for data caching emerges as a game-changing strategy. The low latency, high availability, scalability, and seamless integration make this duo a force to be reckoned with. By adopting Aerospike alongside Snowflake, businesses can ensure their data infrastructure operates at lightning speed, delivering real-time insights and experiences to users, and ultimately accelerating their journey towards success.