Fri. Mar 1st, 2024

Aerospike Inc. today released Version 7 of its real-time, multi-model database with a new unified storage format and other significant in-memory database enhancements.

The unified storage format in Aerospike Database 7 provides the flexibility to choose the right storage engine for different kinds of workloads, even within the same cluster. Developers no longer need to understand the intricacies of in-memory, hybrid-memory, and all-flash storage models. And in-memory deployments gain fast restarts for enterprise-grade resiliency and compression that shrinks the memory footprint of real-time applications.

“Enterprises traditionally turn to in-memory databases for sub-millisecond performance, but they are often brittle, slow to recover, and a hassle to manage and scale,” said Lenley Hensarling, chief product officer, Aerospike. “Now, any enterprise can deploy resilient in-memory applications on a modern, real-time database without compromising on speed, scale, or reliability — and use a fraction of the hardware and development effort of other offerings.”

One-third of Aerospike customers already use Aerospike for in-memory applications. The Aerospike Database optimizes performance with a unique hybrid memory architecture. Enterprises can choose in-memory, all flash, or hybrid storage to maximize performance and scale for a particular workload.

 

Major Aerospike Database 7 New Features

Major features of 7 include:

  • Unified Storage Format for all Storage Engines. A single storage engine and the same programming logic simplifies application deployment and operations.
  • Warm Restarts for High Resiliency. The new in-memory storage engine puts data in shared memory instead of process memory (RAM) enabling restarts in as little as a few seconds — instead of minutes or hours.
  • In-memory Compression Maximizes Memory Efficiency. Version 7 is the first database to enable LZ4, Snappy or ZStandard compression algorithms to gain the same compression ratios regardless of storage engine, saving on hardware cost.
  • In-memory Data Mirrored to Persistence Layer for Best Performance. When deploying in-memory with persistence, all operations like defrag, garbage collection, tomb raiding, and re-balance take place in memory without touching the drive except for mirroring the changes.
  • Deploy without Drives. In-memory namespaces can be deployed without a persistence layer by taking advantage of the Aerospike Shared Memory Tool (ASMT) to persist the namespace from shared memory to the file system after shutdown.

 

One Multi-model Database Simplifies Operations and Reduces Costs

The Aerospike Database handles diverse workloads across the most popular NoSQL data models — key value, document, graph, plus SQL access for analytics — in a single real-time data platform. Aerospike’s multi-model approach simplifies database operations and delivers low-latency, high-throughput processing across data models while handling mixed workloads from gigabyte to petabyte scale.

Operational benefits of Aerospike’s multi-model approach have resulted in proven, best-in-industry total cost of ownership (TCO), which shows that even at the highest levels of scale, Aerospike requires up to 80% less infrastructure than traditional key-value, document, or graph databases.

 

 

By Editor