Active Data Store
Enterprises today need to build personalized experiences through multiple 'web-scale' apps for their 'always-connected' customers. The data from the apps typically moves across three phases in its lifecycle:
- Store: keep-up with fast ingestion needs from users and 'things'
- Analyze: process data & support streaming, batch/ machine learning, interactive querying, etc
- Serve: information (insights, aggregates, etc) at scale and speed
Traditionally, apps have required separate databases and data is moved to data warehouses or lakes across multiple slow processing stages. This cannot meet the real-time expectations of both customers and businesses. Moreover, data is perishable so acting on this active data is critical for businesses to capture revenue opportunities and reduce costs/ risks.
Ampool Active Data Store (ADS) is an in-memory data and compute system that is able to move analytics upstream by providing data management and compute support in an enterprise-grade, scalable, distributed system. By combining multiple specialized in-memory region types in one highly concurrent system, Ampool ADS supports business apps on one side and analytics apps on the other. It allows fast ingest and storage of hot application data, in situ updates and analysis, and data serving from the same scalable distributed data store. As the data ages, Ampool ADS automatically tiers data to warm and cold secondary stores. By speeding the data pipelines several-fold, Ampool enables feeding actionable insights back to applications, driving decisions in a closed loop.
It is suitable for all data processing needs near applications by providing fast access to data across the entire active data lifecycle:
- Store all active data & update it, as required
- Analyze through ‘best-of-breed’ compute engines & frameworks
- Serve data concurrently to multiple data processing stages, tenants & applications
In addition, Ampool supports interfaces to the most common tools used by data 'workers' (data engineers, analysts and scientists) and integrations with large backend batch processing Data Lake's. This allows side by side development of next generation applications that can provide far more context and deeper user experience in near real time.
For further information, please see Ampool Architecture Details