We’re happy to announce the general availability of GigaSpaces XAP 12.1 for download and upgrade.
For the past 15 years, GigaSpaces XAP has been the quintessential in-memory computing platform that powers some of the biggest transaction processing and real-time analytics applications in the world. Today, we’re marking another milestone in this innovation by announcing the release of XAP 12.1.
With XAP 12.1, GigaSpaces brings the power of in-memory computing and combines it with the agility of cloud-native deployments and flexibility of hybrid data storage. XAP 12.1 is an evolution of the existing GigaSpaces In-Memory Computing platform that gives you access to a stable, highly available, and extreme performance environment for all your data and analytics applications. It comes with enhanced capabilities for processing text and unstructured data, frictionless cloud-native deployments for micro-services architectures, and enterprise-ready disaster recovery capabilities through leveraging multi-tiered data storage.
Frictionless Deployment: Focus on the Apps, Not the Administration
XAP 12.1 has been designed to enable developers with continuous deployment and simplified DevOps pipeline by keeping the gap between development and production small. We introduced a RESTful orchestration layer on top of the data grid to provide a frictionless deployment mechanism for microservices development. Regardless of deployment environment, all it takes now for a XAP developer to onboard a new microservice is a couple of REST API calls. The key benefits of this simplified and modern architecture are:
- Leverage declarative “infrastructure as code” patterns for deployment automation to minimize time and cost of on-boarding new applications;
- Deployment parity between on-premises, cloud, and edge obviating the need for customized server administration and monitoring;
- Ability to scale up (increase container sizes) and scale out (adjust container instance count) without any changes to tooling and architecture.
Beyond Structured Data: Search and Text Analytics in Real-Time
Increasingly, businesses are finding themselves in a world where most of their customers are using social media to talk about their experiences. Unlike structured transactional data (shopping carts items, call data records), customer feedback is all in the form of free text, just waiting to be gathered and analyzed. Whether it’s a product review or a giant Reddit thread about your company, engaging customers means analyzing their textual feedback in real-time. Traditionally, text search engines (such as Apache Lucene/Solr) are implemented as infrastructure separated from the main applications. While this works for performing historical reports, reacting to customer feedback in real-time requires a co-location of the search capabilities with the rest of customer transactions and history. This fulfills the elusive quest towards customer-360 and hyper-personalization for the omnichannel digital enterprise.
XAP 12.1 introduces Full-Text Search API based on Lucene indexes and analyzers so users can run search queries (wildcard, fuzzy match) in memory at high throughput and low latencies. Combined with the rest of XAP’s event-driven container API, applications can trigger events and messages in the moment based on a text search criteria.
Hybrid Storage Engine, Record-Fast Data Recovery
Traditionally, one of the main drawbacks to in-memory computing is the ephemeral nature of RAM. In case of a planned or unplanned grid restarts, all the data has to be reloaded into memory. When bringing an in-memory data grid online after an outage, two concerns must be addressed:
- Making sure all the data held at the moment prior to downtime persists;
- Getting all the data back online as fast as possible to ensure business continuity.
The business impact of failing to deliver on both can be financially devastating. While the first concern has already been addressed by GigaSpaces since XAP 10.1 release, the second one required manual data re-ingestion processes to be put in place. XAP 12.1 makes the “cache warm-up” problem a thing of the past. With XAP 12.1’s MemoryXtend Initial Load feature, customers are now able to specify (through a SQL query) a set of data that immediately gets reloaded into memory upon restart. This not only improves the recovery time objective in disaster recovery scenarios but also immediately warms the data grid with critical data that require low latency access.