Telegraf agent plugin activation via Suite API fails with HTTP 422 "Configuration not valid"

Products

• VCF Operations

Issue/Introduction

Activating a plugin (for example, Processes / procstat) through the Suite API fails consistently.

• The API returns an HTTP 422 error with the message: "Configuration not valid."

• In Recent Tasks, the job shows a status of "Not Dispatched" with the error: "Failed to retrieve required details for resource."

Note: Performing the exact same action from the UI (opening the configuration, editing, and saving) works without issues.

Environment

VCF Operations 9.1

Cause

Operations uses a property called UCPFQDN to identify which Cloud Proxy manages a monitored machine. When a plugin action is triggered, Operations reads this property to route the work correctly.

This failure occurs due to a combination of three factors:

1. Broken Machine in Inventory: There is an Endpoint object in the cluster with an empty UCPFQDN (it has no agent installed, so this property is never populated).

2. Accidental Inclusion of All Endpoints: Due to a known issue, activating a plugin via the API on a single machine accidentally attaches every Endpoint in the entire cluster to the request instead of just the target machine.

3. Strict Validation Checks: Operations verifies that it can find a Cloud Proxy for every attached machine before executing the job. If even one machine (like the broken endpoint mentioned above) fails this check, the entire request is canceled. There is no mechanism to skip the broken endpoint and proceed.

This explains why the UI succeeds: the UI uses a different code path that correctly scopes the request to only the specifically selected machine.

Resolution

Broadcom is aware of this issue issue and it will be fixed in upcoming release of VCF 9.1

Workaround:

1. Log in to the VCF Operations UI, expand Workloads from the left navigation menu, and click on Applications.
2. Click the Manage Telegraf Agents tab and filter for endpoints that are in an "Agent not installed" state.
3. Select the affected endpoint, view its Properties, and check if the UCPFQDN property is empty.
4. If the UCPFQDN property is empty, go to your Inventory and delete that specific endpoint resource.
5. Retry activating the plugin on your intended object via the VCF Operations API.

Example: