canonical: https://jentic.com/apis/googleapis.com/google-memorystore-memcached

# Google Cloud Memorystore for Memcached API

Cloud Memorystore for Memcached provides a managed Memcached service on Google Cloud, with this API covering instance lifecycle, parameter management, version upgrades, and maintenance scheduling. Instances are scoped to project locations and most state changes return long-running operation handles. The API exposes apply-parameters and update-parameters semantics that let teams change Memcached configuration on running nodes, and rescheduleMaintenance lets workloads control planned maintenance windows.

## For AI agents

Provision and manage Cloud Memorystore for Memcached instances: create and delete instances, change Memcached parameters on live nodes, upgrade engine versions, and reschedule maintenance.

## Scope

Does not handle cache reads or writes, persistence, or Redis-style data structures - use for Memcached instance lifecycle, parameter management, version upgrades, and maintenance scheduling only.

## Capabilities

- Provision and delete Memcached instances inside a Google Cloud project location
- List Memcached instances and their node counts and versions
- Update Memcached parameters and apply them across the running nodes
- Trigger an in-place version upgrade of a Memcached instance
- Reschedule a planned maintenance window for an instance
- Cancel and inspect long-running provisioning and update operations

## Use cases

### Cache layer provisioning

Provision a managed Memcached instance to sit in front of a backend database for read-heavy workloads, specifying node count, node CPU and memory, and the network. Provisioning is a long-running operation that typically completes within a few minutes for small instances.

Example prompt: Create an instance named cache-prod under projects/acme/locations/us-central1 with 3 nodes of 4 vCPUs and 4 GB memory each and poll the operation until done

### Live parameter tuning

Update Memcached parameters such as max_item_size on a running instance and use applyParameters to push the change to all nodes without recreating the cluster. Useful when applications begin storing larger objects or require connection-related tuning.

Example prompt: Update parameters on instance cache-prod to set max_item_size=4194304, then call applyParameters on all node IDs

### Maintenance window control

Reschedule a planned maintenance window so it falls outside critical traffic periods. The rescheduleMaintenance endpoint accepts a target start time or instructs the service to take the next available window, returning a long-running operation that resolves once the schedule is updated.

Example prompt: Call rescheduleMaintenance on instance cache-prod with rescheduleType SPECIFIC_TIME and scheduleTime set to next Sunday 03:00 UTC

### Agent integration via Jentic

AI agents managing cache infrastructure can use the Memcached API through Jentic to provision instances, tune parameters, and time maintenance windows without holding the underlying OAuth refresh token. Jentic surfaces the long-running operation handles for the agent to poll.

Example prompt: List all instances in projects/acme/locations/us-central1 with no traffic in the last 7 days and queue a delete request for each

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | /v1/{+parent}/instances | List Memcached instances in a project location |
| DELETE | /v1/{+name} | Delete a Memcached instance |
| POST | /v1/{+name}:applyParameters | Apply staged parameters to nodes |
| PATCH | /v1/{+name}:updateParameters | Update Memcached parameters |
| POST | /v1/{+name}:upgrade | Upgrade a Memcached instance version |
| POST | /v1/{+instance}:rescheduleMaintenance | Reschedule a maintenance window |
| GET | /v1/{+name}/operations | List long-running operations |

## Key resources

- **instances** — Provision, list, update, delete, upgrade, and apply parameters on Memcached instances
- **operations** — Track and cancel long-running operations for instance lifecycle calls
- **locations** — Enumerate the Google Cloud locations where Memcached can be provisioned

## Why Jentic

- **Setup:** Wiring the Cloud Memorystore for Memcached API by hand means setting up Google OAuth 2.0 with the cloud-platform scope, minting bearer tokens, and polling long-running instance operations yourself against the memcache.googleapis.com host. Through Jentic you install once, import the Cloud Memorystore for Memcached API from the API Directory, store the OAuth credential once, and your agent calls it.
- **Permission scoping:** The Memcached API puts the instance name in the URL path (/v1/{+name}...), so a rule can pin your agent to one Memcached instance: it can apply and update parameters there and nothing else. You choose the operations it may call, so destructive ones like instance deletion or a version upgrade are not included unless you add them.
- **Credential handling:** Your Google OAuth credential is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'provision a Memcached instance' or 'update Memcached parameters', and Jentic returns the matching Memorystore operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

## Related APIs

- **Cloud Memorystore for Redis API** — Managed Redis on Google Cloud for richer data structures and persistence
- **Compute Engine API** — Provisions VMs that connect to the Memcached instance over the same VPC network
- **Cloud Monitoring API** — Reads time-series metrics emitted by Memcached instances for alerting and dashboards

## FAQ

### What authentication does the Memorystore for Memcached API use?

The API uses Google OAuth 2.0 with the https://www.googleapis.com/auth/cloud-platform scope. Through Jentic, the OAuth credentials are held in the encrypted vault and the agent only sees a scoped, short-lived bearer token.

### Can I update Memcached parameters on a running instance?

Yes. Call updateParameters with the new parameter values, then applyParameters to push the staged changes to nodes. Both calls return long-running operations you can poll for completion.

### What are the rate limits for this API?

Per-project quotas for the Memorystore service govern how many instances you can have and how often lifecycle operations can be invoked. See the Cloud Console quotas page; bulk lifecycle calls should respect concurrent operation limits and use exponential backoff on 429 responses.

### How do I reschedule a Memcached maintenance window through Jentic?

Search Jentic for reschedule memcached maintenance, load the operation that maps to /v1/{+instance}:rescheduleMaintenance, and execute with the instance name and a SPECIFIC_TIME scheduleTime. Jentic returns the long-running operation for polling.

### Is the Memorystore for Memcached API free?

The API itself is not separately billed; you pay for provisioned Memcached node-hours and any associated network egress. See the Memorystore pricing page for current rates.

### Does this API let me read or write cache values?

No, this is the control-plane API for managing instances. Application-level cache reads and writes happen over the standard Memcached protocol against the instance's discoveryEndpoint, not through this API.

### Can I limit what my agent is allowed to do with the Cloud Memorystore for Memcached API?

Yes. Because your Jentic One instance is self-hosted, your own rules decide which operations and credentials the agent may use. Since the Memcached API carries the instance name in the URL path, you can pin the agent to a single instance and grant it only the calls you want, such as updateParameters and applyParameters. Destructive operations like deleting an instance or triggering a version upgrade stay off limits unless you explicitly add them.