Intelligent Data Tiering

As the data requirements for your applications increase, you may reach a point where you must make a smart decision how to tier your data storage -- which data should be in rapid-access RAM, and which data can be stored in lower cost media.

GigaSpaces provides several ways to manage the automatic data life cycle for multiple storage tiers, so you can balance your storage media costs and required performance for different data that you wish to store and process.

How Do We Decide Which Data is Most Important?

This decision — which data must be in RAM and which data can reside in lower tiers — can be based on several criteria:

• Required SLA

• Frequency of access/data aging

• Frequency of change (mutability)

• Relative importance for business purposes

GigaSpaces provides several tiering solutions that provide a balance between performance and cost of operations:

• What is Tiered Storage? — Automatically tier your data between RAM (Hot tier) and SSD (Warm tier Part of GigaSpaces Tiered Storage Mechanism. Warm tier is SSD (recommended media) storage which is mostly used for data that is read-only and is less frequently used.) according to business rules, or based on time. For example, data may be placed in the Hot tier for premium customers, or if the data has been accessed within the previous two weeks.
  Note that data in the Hot tier is highly mutable and supports both operational and analytical workloads, while data in the Warm tier is primarily for read access or for use in analytical workloads.

• MemoryXtend — Tier RAM and SSD, but keep all indexes in RAM for better performance. MemoryXtend Related to Data Tiering. The MemoryXtend (blobstore) storage model allows an external storage medium (one that does not reside on the JVM heap) to store the GigaSpaces Space data and is designed for operational workloads. It keeps all indexes in RAM for better performance. is designed for operational workloads.

• AnalyticsXtreme — Real-time data analytics over a data lake of real-time and historical data