Kyvernon

Reference For: administrators, users, developers

Performance data errors

Causes and remedies for the error codes of performance charts, metric queries and sample ingestion.

error.metrics_disabled

Text in the console
Performance data is turned off. The management service runs with performance metrics disabled. Ask an administrator to enable them in the controller configuration.
Meaning for API clients
Performance metrics are turned off on the controller.
Code
UNAVAILABLE, HTTP 503

Cause. Performance metrics are turned off on the controller, so charts and the usage ledger have no data to serve. The setting is performance in the [metrics] section of the controller configuration.

Remedy.

  1. An administrator sets performance = true in the [metrics] section of the controller configuration and restarts the controller. The value is true by default, so the setting was changed on purpose.
  2. Pass on the message id error.metrics_disabled when you cannot change the configuration yourself.

error.metrics_bad_entity

Text in the console
Performance data is not kept for this object. Only hosts, virtual machines and datastores have charts. Open one of those objects.
Meaning for API clients
No performance data is kept for that object kind.
Code
INVALID_ARGUMENT, HTTP 400
ArgumentWhat it holds
entitythe id that was rejected.

Cause. The request asks for performance data of an object kind that has none. Only hosts, virtual machines and datastores have performance data.

Remedy.

  1. Open a host, a virtual machine or a datastore and use its performance view.
  2. With the CLI, give the id of one of those objects: kyvernon metrics available <entity>.

error.metrics_bad_interval

Text in the console
The time range is not available. The management service keeps 20 s, 5 min, 30 min, 2 h and 1 day samples. Choose one of the listed ranges.
Meaning for API clients
Not a supported sampling period.
Code
INVALID_ARGUMENT, HTTP 400
ArgumentWhat it holds
intervalthe period in seconds that was rejected.

Cause. The sampling period is not one that Kyvernon keeps. The periods are 20 seconds, 5 minutes, 30 minutes, 2 hours and 1 day.

Remedy.

  1. Choose one of the listed periods. In the console, pick a time range with Time range.
  2. With the CLI, list the periods: kyvernon metrics intervals.

error.metrics_unknown_counter

Text in the console
A requested counter does not exist. The management service may run an older version. Update the management service or reload the page.
Meaning for API clients
No such counter.
Code
INVALID_ARGUMENT, HTTP 400
ArgumentWhat it holds
counterthe name or id that was rejected.

Cause. The request names a counter that the controller does not know. The name has a typo, or the console or script is newer than the controller.

Remedy.

  1. List the counters that the controller knows: kyvernon metrics counters.
  2. Use a name or an id from that list.
  3. When the console is newer than the controller, update the controller, or reload the page.

error.metrics_query_empty

Text in the console
The request named no object or counter. Reload the page and try again.
Meaning for API clients
The query names no entity or counter.
Code
INVALID_ARGUMENT, HTTP 400

Cause. The query names no object, or no counter. A query needs at least one of each.

Remedy.

  1. Give at least one entity and one counter: kyvernon metrics query --entity <id> --counter <name>.
  2. In the console, reload the page and try again.

error.metrics_query_too_large

Text in the console
The request covers too many samples. A query returns at most 200,000 points. Choose a shorter time range or fewer devices.
Meaning for API clients
The query would return too many samples.
Code
INVALID_ARGUMENT, HTTP 400

Cause. The query asks for too much. A query names at most 100 entities and 100 counters, and returns at most 200,000 samples.

Remedy.

  1. Choose a shorter time range, a longer sampling period or fewer objects. kyvernon metrics query accepts --start, --end, --interval and --max-samples.
  2. Split the query into several smaller ones.

error.metrics_too_many_batches

Text in the console
A host agent sent more than {max} metric batches in one report, so the whole report was rejected. Update the host agent, or contact support if the agent is current.
Meaning for API clients
A host agent sent more than max metric batches in one report; the whole report is rejected.
Code
INVALID_ARGUMENT, HTTP 400
ArgumentWhat it holds
maxthe most batches the controller accepts in one report.

Cause. A host agent sent more than max sample batches (16) in one report, so the controller rejected the whole report. Nothing is wrong with the metrics service. Console users do not cause this error; it concerns the connection between a host agent and the controller.

Remedy.

  1. Update the host agent to the version of the controller. Both come from the same release.
  2. When it stays, an administrator reads the logs of the controller and the host agent. Pass on the message id and the name of the host.

Version 0.1. Last changed: 2026-11-13