MCP Server (Preview Mode)

In version 17.2.2 we introduce the MCP Server, bringing native Model Context Protocol (MCP) support to GigaSpaces clusters. Currently in Preview Mode, this feature allows customers to expose XAP GigaSpaces eXtreme Application Platform. Provides a powerful solution for data processing, launching, and running digital services's ultra-low-latency data tier and distributed compute capabilities directly to autonomous AI agents and agentic workflows.

To protect grid environments from aggressive or runaway AI agent behavior, every deployed MCP service includes configurable Rate Limiting (operations per minute) and utilizes Service Account tokens to enforce strict role-based access control.

The Agentic Compute & Execution Engine

Instead of moving massive enterprise datasets across the network to an LLM—which wastes critical time and API tokens—you can now bring the AI's code instructions directly to the data.

AI agents can leverage a suite of dedicated User-Defined Logic & Compute Tools to trigger high-performance processing natively inside the cluster.

Modular MCP Configurations & Flexible System Capabilities

Administrators can instantiate multiple, independent MCP servers targeted at different business needs. Through the user interface, you can precisely configure each service by selecting only the specific functional toolsets required for that target, allowing you to define grid administrator agent, various logic running agents and agents who change data based on insights from other flows. See more at XAP MCP Tools Overview

Semantic Layer

The Semantic Layer serves as the foundational bridge between Large Language Models (LLMs) and the XAP data model, automatically translating complex cluster schemas into clear, business-friendly concepts that AI agents can effortlessly navigate.

The entire Semantic Layer is hosted on a specialized, dedicated space configured with a META mode. It includes Unified Cluster Metadata & Relationships and User-Defined Tool & Parameter Descriptions. To further empower agentic decision-making, the Semantic Layer will be enhanced to actively track and expose real-time statistics on each table in future versions.

See more at Semantic Layer Foundation

Vector Search

Vector Search is an XAP extension that enables semantic similarity search on high-dimensional vector embeddings. Instead of exact keyword matching, you can find conceptually similar items based on their vector representations. See more at Vector Search.

We value your input! As these capabilities are in preview, we want to hear about your unique use cases. Please contact us to collaborate on optimizing your agentic workflows or if your system requires any additional capabilities.

Redolog drop packet listener

Added ability to listen to drop packets in the embedded space see: Dropped Packet Listener

Release Date: March-2026

This page describes product changes, along with new features and functionality for GigaSpaces release 17.2.2.

Next-Gen Management: REST API V3

The V3 Manager API is a complete modernization of our management interface, built on the OpenAPI 3.1.0 specification.

The configuration is located at gs-home\config\ui\rest-v3.yaml.

The default port is 9090. This can be changed by the environment variable GS_REST REpresentational State Transfer. Application Programming Interface An API, or application programming interface, is a set of rules that define how applications or devices can connect to and communicate with each other. A REST API is an API that conforms to the design principles of the REST, or representational state transfer architectural style._V3_PORT.

Swagger URL: http://<manager>:9090/api/v3/swagger-ui/index.html#

See more at REST Manager API.

Key API Enhancements

• Processing Units This is the unit of packaging and deployment in the GigaSpaces Data Grid, and is essentially the main GigaSpaces service. The Processing Unit (PU) itself is typically deployed onto the Service Grid. When a Processing Unit is deployed, a Processing Unit instance is the actual runtime entity. (PUs): Expanded endpoints for vertical and horizontal scaling, unquiescing, detailed instance utilization tracking, event container definitions and statistics, and remote services.

• Space Where GigaSpaces data is stored. It is the logical cache that holds data objects in memory and might also hold them in layered in tiering. Data is hosted from multiple SoRs, consolidated as a unified data model. Intelligence: New endpoints for space utilization, durable task management, and detailed transaction monitoring, including the ability to identify specific locked objects, details about local cache and local views, registration of new types, generating demo data, and more.

• Monitoring Control: Granular control over LRMI monitoring (Low-level Remote Method Invocation) can now be toggled per space via the API, allowing traffic monitoring per method and listing of recent alerts.

• Host and Container Data: Enhanced "Extended" endpoints provide deeper operating system details and utilization metrics for physical hosts and their associated containers.

Modernized Metrics (Micrometer.io)

We have transitioned from our proprietary metrics framework to Micrometer.io, the industry standard for JVM Java Virtual Machine. A virtual machine that enables a computer to run Java programs as well as programs written in other languages that are also compiled to Java bytecode.-based monitoring. This change ensures seamless integration with a wide range of monitoring ecosystems.

Configuration and Management

For more information, see Configuration.

• Property-Based Configuration: Metric configuration has shifted from XML to a centralized properties file located at gs-home\config\metrics\metrics.properties.

• InfluxDB Integration: The new V3 API includes a dedicated endpoint (POST /api/v3/metricsinfluxdb) to execute complex InfluxDB metrics queries directly through the manager to ensure backward compatibility.

Grafana Dashboard Migration

Existing dashboards are updated to the new Micrometer-compatible versions (the old ones are in the micrometer/ refs). Key changes include:

• Metric Renaming: CPU usage has moved from os_cpu_used-percent (0–100) to system_cpu_usage (0.0–1.0).

• Throughput Logic: Monotonically increasing cumulative counters replace pre-computed throughput. Queries now use the non_negative_derivative(..., 1s) function to calculate operations-per-second.

• Tagging Updates: Template variables for object type drilldowns now use the space_name tag instead of pu This is the unit of packaging and deployment in the GigaSpaces Data Grid, and is essentially the main GigaSpaces service. The Processing Unit (PU) itself is typically deployed onto the Service Grid. When a Processing Unit is deployed, a Processing Unit instance is the actual runtime entity._name.

Enhanced Observability (OpenTelemetry)

GigaSpaces now includes a built-in OpenTelemetry SDK (v1.40.0) for distributed tracing, exporting spans to Zipkin-compatible backends. For more information, see Distributed Tracing.

• Automatic Instrumentation: Child spans are automatically generated for core GigaSpace operations, including read, write, take, change, aggregate, and execute.

• Zero Overhead Tracing: The framework uses "opt-in" tracing. Spans are only created if a valid parent context exists in the current thread, ensuring no performance impact on untraced requests.

• Context Propagation: W3C Trace Context (traceparent / tracestate) is supported by default for correlation across multi-service environments.

Unified Control: SpaceDeck UI Integration

SpaceDeck GigaSpaces intuitive, streamlined user interface to set up, manage and control their environment. Using SpaceDeck, users can define the tools to bring legacy System of Record (SoR) databases into the in-memory data grid that is the core of the GigaSpaces system., our premier data visualization and management interface, is now natively integrated into the GigaSpaces Manager. Configuration is located at gs-home\config\ui\spacedeck.yaml.

• Centralized Access: Manage the grid, explore data, and monitor performance through a single, unified interface.

• Environment Awareness: You can now visually distinguish environments by setting labels and background colors via environment variables:

  • SPACEDECK_ENV_LABEL (e.g., "PROD", "STAGING")

  • SPACEDECK_ENV_BG_COLOR (Hex color codes, e.g., #FF0000 for red)

Security Note: CVE-2022-39135

The vulnerability involves XML External Entity (XXE) injection via specific Calcite functions. This is not a risk in GigaSpaces for the following reasons:

• Disabled Functions: The specific functions and XML parsing features associated with this vulnerability are not utilized or exposed within the GigaSpaces implementation of Calcite.

• Sanitized Input: All queries processed by the Calcite engine in GigaSpaces are strictly validated and handled as data-centric operations on the grid, preventing the execution of arbitrary external entity lookups.

We continue to track industry standards and ensure our dependency management aligns with the latest security patches to maintain a robust and secure environment.