Cenovel / Operate

Health & operations

Read device observations accurately and check the appliance’s own health.

Device state is an observation

When a read cannot reach a device, the console describes it as not answering Cenovel. A rejected credential or untrusted key is distinguished from a device that does not answer.

When attempted device reads fail together and nothing has answered recently, Cenovel raises Cenovel can’t reach the management network. While that condition holds, device reachability outages are held: no new one opens, existing ones are not confirmed or cleared, and their clocks stop.

Analytics

Analytics uses one query engine across Answers, Explore, and Saved. Only Administrator EX can share or schedule views. Department scope and page permissions are enforced by the server.

Everyday appliance checks

cenovelctl status
cenovelctl health
cenovelctl diagnose

Status and diagnose accept --json. Diagnose exits non-zero when its health checks fail.

Inspect logs

cenovelctl logs api --lines 200
cenovelctl logs worker --follow
cenovelctl support-bundle --json

The support-bundle command prepares redacted diagnostics. Review the bundle before sharing it.

Liveness, readiness, and metrics

  • /livez: the process answers.
  • /readyz: lifecycle readiness and a database query pass; when daily application backups are enabled, their subsystem must be available. Failure returns 503.
  • /metrics: Prometheus-format metrics, available to a loopback request or a request with the configured bearer token.

Readiness does not prove an off-box backup or a successful restore. Check those results separately.

Read the operator reference
5. Observations, health, and diagnostics

A device that cannot be reached by a read is described as "not answering
Cenovel". A rejected sign-in or untrusted key is distinguished from no answer.

When attempted reads fail together and nothing has answered recently,
Cenovel raises "Cenovel can't reach the management network". While that
condition holds, device reachability outages are held: no new one opens,
existing ones are neither confirmed nor cleared, and their clocks stop.

Native appliance checks:

    cenovelctl status
    cenovelctl diagnose
    cenovelctl health
    cenovelctl logs api --lines 200
    cenovelctl logs worker --follow
    cenovelctl support-bundle --json

Status and diagnose support --json. Diagnose exits non-zero when its health
checks fail. The support bundle contains redacted diagnostics; review it
before sharing. The console also sends performance measurements to its own
local API. These measurements support local diagnostics.

/livez answers when the process responds. /readyz checks lifecycle readiness
and a database query; when daily application backups are enabled, the backup
subsystem must be available. Failures return 503. /healthz is an alias.
Readiness does not prove an off-box backup or a successful restore.

/metrics returns Prometheus-format metrics to a loopback request or a request
with the configured bearer token. Other requests receive 404.

Reviewed against the current source code. Native appliance administration commands require the appropriate host privileges; validate deployment procedures on your own release before production use.