Working together to accelerate software-defined storage adoption across the storage industry.
Store, process, and analyze data-heavy apps in real time with a smaller RAM footprint
The GigaSpaces XAP In-Memory Computing Platform delivers fast processing of all your data and applications in real time. With the XAP MemoryXtend add-on, optimized access to any SSD enables a hybrid storage model in which hot data is stored on RAM and raw data is stored on an external solid-state drive. XAP enables an entire application to run on a single platform with all tiers collapsed into one container, while storing all data in memory, and scaling applications automatically and on demand. It allows for extreme data management using in-memory data caching, while ensuring immediate consistency, fully supporting transactions and rich query semantics and data indexing.
XAP is optimized for big data by efficiently combining real-time processing of massive data streams with batch processing of large data volumes using built-in integrations (e.g., Hadoop* or Cassandra*). Notably, XAP’s cross-data center scalable replication functionality enables near real-time replication of massive data streams, which is essential for disaster recovery, cloud bursting, and multisite scenarios.
Hundreds of Tier 1 organizations leverage XAP to enhance IT efficiency and performance, including top financial firms, e-commerce companies, online gaming providers, healthcare organizations, and telecom carriers.
By default, XAP entries are stored in-memory to provide the fastest performance possible. However, as data grows in size and numbers and modern storage technologies such as SSD and 3D XPoint™ provide faster data access, customers are looking for smarter solutions to get the best of both worlds.
This is where XAP MemoryXtend comes into play. XAP MemoryXtend offers a hybrid storage model, which stores indexes in-memory and data off-heap (e.g., SSD, or 3D XPoint), so that the data grid size grows significantly, and performance impact is minimal, since most of the query’s work is done using the in-memory index.
The current MemoryXtend plug-in is file system-based, allowing it to be used as-is on a 3D XPoint-based device mounted to the system. However, the MemoryXtend architecture is designed for extensibility, so once the 3D XPoint community builds an SDK providing optimized access, XAP will leverage it to implement a new MemoryXtend plug-in, providing even better performance.
XAP MemoryXtend—Massive Application Storage Capacity for Real-Time Applications
This white paper reviews a potential solution for the issue of prohibitive costs associated with in-memory computing. XAP MemoryXtend enables a hybrid architecture which intelligently stores data on both RAM and SSD, allowing its users to enjoy both high-speed processing and cheaper storage capacity. The paper presents a benchmark that weighs XAP MemoryXtend’s price against performance and shows that it is indeed the optimal solution in this context.
Customer Success Story: Supporting Data Growth and Faster Processing of Mission-Critical Data with GigaSpaces XAP MemoryXtend
Sentri7*, Wolters Kluwer Health’s flagship patient-surveillance offering, uses a high-performance rules engine combined with an in-memory data grid based on XAP for surveillance, intervention, documentation, and reporting. Whenever IT maintenance on Sentri7 is required, it is necessary to restart the grids, often taking anywhere from 30 minutes to over an hour to complete the process. With GigaSpaces XAP MemoryXtend, this process can be done 20 times faster.