For Agents
Operate an Infinispan distributed cache cluster: create caches, read/write entries, manage counters, control cross-site replication, run rolling upgrades, and administer the cluster.
Use for: I want to create a new distributed cache called 'sessions', Put a value into a cache by key, Retrieve an entry from the 'orders' cache, List all caches running on the Infinispan cluster
Not supported: Does not handle managed cloud database hosting, persistent SQL queries, or full-text relevance ranking - use for distributed cache, in-memory data grid, atomic counters, and cluster administration only.
Infinispan is an open-source distributed in-memory key/value data store, also used as a clustered cache and embedded data grid. The 94-endpoint REST API exposes cache lifecycle and entry CRUD, atomic counters, cross-site replication controls, search and indexing, Protobuf schema management, server and cluster administration, security, rolling upgrades, and backups. The base URL is operator-supplied because Infinispan is typically self-hosted; authentication is HTTP Basic or Digest depending on server configuration.
Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Infinispan REST API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Finfinispan.org%2Finfinispan" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Finfinispan.org%2Finfinispan" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with Infinispan REST API.
Create, modify, clear, and delete named caches via /rest/v2/caches/{cacheName} including alias assignment and rebalancing controls
Add, replace, retrieve, and delete cache entries with /rest/v2/caches/{cacheName}/{cacheKey}
Convert and compare cache configurations across templates and live caches without taking the cluster offline
Manage atomic counters under /rest/v2/counters with strong and weak consistency semantics
GET STARTED
Configure cross-site replication: take sites offline, push state to remote sites, and inspect replication status
Run rolling upgrades by routing traffic between source and target clusters during version transitions
Manage Protobuf schemas, indexed search queries, and server-level security and configuration
Patterns agents use Infinispan REST API for, with concrete tasks.
★ Distributed Session and Application Cache
Operate Infinispan as a session store or hot-data cache in front of a slower system of record. POST /rest/v2/caches/{cacheName} creates the cache from a template, POST /rest/v2/caches/{cacheName}/{cacheKey} writes entries, and GET /rest/v2/caches/{cacheName}/{cacheKey} reads them back. Suitable for stateless application servers that need a shared cache layer with replication and failover.
Create a cache named 'sessions' with a distributed-sync template, then POST a JSON value to /rest/v2/caches/sessions/user-123.
Cross-Site Replication and Disaster Recovery
Configure and operate Infinispan's cross-site replication, where a primary datacenter pushes cache state to one or more remote sites. The /rest/v2/cache-managers/{name}/x-site endpoints expose actions to take a remote site online or offline, push state, and inspect replication backlog. Critical for active-active deployments and DR drills where one site has to take over from another within seconds.
Take the 'NYC' remote site offline for a planned datacenter migration and confirm replication backlog clears on the source side.
Rolling Upgrade Between Infinispan Versions
Migrate an Infinispan cluster from one major version to another without downtime by running source and target clusters concurrently and routing traffic between them. The /rest/v2/cluster?action=connect-source and related rolling-upgrade endpoints orchestrate the cutover. Reduces planned-maintenance windows for caches that back live customer-facing systems.
Initiate a rolling upgrade from cluster A to cluster B, monitor the state transfer, and disconnect once the target reports a complete state.
AI Agent Operating a Cache Cluster
An ops agent administers Infinispan deployments - provisioning caches, clearing stale data, taking sites offline during incidents - through Jentic. The agent searches by intent (e.g., 'clear a cache'), loads the schema, and executes against the cluster. Useful for SRE copilots that handle cache lifecycle as part of broader incident playbooks.
List all caches with GET /rest/v2/caches, identify those with zero entries, and POST /rest/v2/caches/{cacheName}?action=clear on each to free memory.
94 endpoints — infinispan is an open-source distributed in-memory key/value data store, also used as a clustered cache and embedded data grid.
METHOD
PATH
DESCRIPTION
/rest/v2/caches/{cacheName}
Create a cache from a template or configuration
/rest/v2/caches/{cacheName}
Get cache details and configuration
/rest/v2/caches/{cacheName}
Delete a cache
/rest/v2/caches/{cacheName}/{cacheKey}
Add a cache entry
/rest/v2/caches/{cacheName}/{cacheKey}
Retrieve a cache entry by key
/rest/v2/caches/{cacheName}?action=clear
Clear all entries in a cache
/rest/v2/caches?action=compare
Compare two cache configurations
/rest/v2/caches/{cacheName}?action=enable-rebalancing
Enable cluster rebalancing for a cache
/rest/v2/caches/{cacheName}
Create a cache from a template or configuration
/rest/v2/caches/{cacheName}
Get cache details and configuration
/rest/v2/caches/{cacheName}
Delete a cache
/rest/v2/caches/{cacheName}/{cacheKey}
Add a cache entry
/rest/v2/caches/{cacheName}/{cacheKey}
Retrieve a cache entry by key
/rest/v2/caches/{cacheName}?action=clear
Clear all entries in a cache
/rest/v2/caches?action=compare
Compare two cache configurations
/rest/v2/caches/{cacheName}?action=enable-rebalancing
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the Infinispan REST API by hand means configuring Basic or Digest auth, supplying your own cluster host and port, and mapping its cache and admin operations yourself. Through Jentic you install once, import the Infinispan REST API from the API Directory, store the credential once, and your agent calls it.
Permission scoping
Infinispan puts the cache name in the URL path (/rest/v2/caches/{cacheName}), so a rule can pin your agent to one cache: it can read and write entries there and nothing else. You choose the operations it may call, so destructive ones like clearing a cache or taking a site offline are not included unless you add them.
Credential isolation
Your Infinispan Basic or Digest credentials are stored once, encrypted, by your own Jentic One instance and injected at execution time. They never enter the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'create a cache' or 'clear a cache', and Jentic returns the matching Infinispan operation with its input schema so the agent calls the right endpoint without navigating the 94-endpoint admin surface.
Alternatives and complements available in the Jentic catalogue.
Specific to using Infinispan REST API through Jentic.
What authentication does the Infinispan REST API use?
The Infinispan REST API supports HTTP Basic and HTTP Digest authentication, configurable on the server. Through Jentic, the credentials for the cluster endpoint are stored encrypted in your Jentic One instance and injected at runtime, so the agent never holds the raw username/password pair.
Can I create caches and write entries through the REST API?
Yes. POST /rest/v2/caches/{cacheName} creates a cache from a configuration body or named template, and POST /rest/v2/caches/{cacheName}/{cacheKey} writes an entry. GET /rest/v2/caches/{cacheName}/{cacheKey} reads it back, and DELETE removes it.
What are the rate limits for the Infinispan REST API?
Infinispan is self-hosted, so there are no platform-imposed rate limits - throughput is bounded by the cluster size, network, and cache configuration. Server-side ACL roles and request quotas can be configured if you need to throttle specific principals.
How do I clear a cache through Jentic?
Search Jentic for 'clear an Infinispan cache' to find POST /rest/v2/caches/{cacheName}?action=clear. Load the schema, supply cacheName, and execute. The action removes all entries without deleting the cache itself. Get started with Jentic One, the self-hosted execution layer.
Does the API support cross-site replication operations?
Yes. The /rest/v2/cache-managers/{name}/x-site and related endpoints let you take remote sites online/offline, push state, and inspect replication backlog - useful for active-active deployments and DR drills.
Can I run a zero-downtime upgrade between Infinispan versions?
Yes. The rolling-upgrade endpoints (/rest/v2/cluster?action=connect-source and friends) orchestrate running source and target clusters concurrently while traffic migrates, so the cache stays available during the version cutover.
Can I limit what my agent is allowed to do with the Infinispan REST API?
Yes. Because you run Jentic One yourself, your own rules decide which Infinispan operations and credentials the agent may use. Since the cache name sits in the URL path (/rest/v2/caches/{cacheName}), you can pin the agent to a single cache so it only reads and writes entries there. You also pick the exact operations it may call, so destructive ones like clearing a cache or taking a cross-site replica offline stay out of reach unless you add them.
Enable cluster rebalancing for a cache