Technologies/Prefect/prefect.agent.heartbeat
PrefectPrefectMetric

prefect.agent.heartbeat

Agent heartbeat freshness
Dimensions:None
Available on:Native (1)
Interface Metrics (1)
Native
Time since last agent heartbeat in seconds
Dimensions:None
Sources
agent_heartbeat_seconds_agodocs.prefect.io

Technical Annotations (44)

Configuration Parameters (1)
checkpoint_everyrecommended: 1
ensures EventsClient sends events immediately rather than buffering
Error Signatures (8)
Invalid input ConnectionInputs.SEND_HEADERS in state ConnectionState.CLOSEDexception
h2.exceptions.ProtocolErrorexception
httpcore.LocalProtocolErrorexception
httpx.LocalProtocolErrorexception
Agent stopped!log pattern
Failed the last 3 attemptslog pattern
FileNotFoundError: [Errno 2] No such file or directory: 'prefect'exception
cannot find a heartbeaterror code
CLI Commands (4)
prefect agent start --pool default-agent-pool --work-queue defaultdiagnostic
python -m sitediagnostic
. /path/to/.venv/bin/activate && /path/to/.venv/bin/dask-workerremediation
prefect agent startdiagnostic
Technical References (31)
runnercomponentghost runsconceptworkercomponentRunnercomponentEventsClientcomponentEventWorkercomponenth2componenthttpcorecomponenthttpxcomponentHTTP/2protocolcritical_service_loopcomponentcheck_for_cancelled_flow_runscomponentget_and_submit_flow_runscomponentsubprocess.PopencomponentDaskExecutorcomponentsystemdcomponent/prefect/utilities/executors.pyfile pathsupervisordcomponentdocker-in-dockerconceptscreencomponentagentcomponentflow runconceptnotificationcomponentlate stateconceptk8s agentcomponentliveness probeconceptPrefect CloudcomponentZombie Killer Servicecomponentwork queuecomponentKarpentercomponentspot instanceconcept
Related Insights (13)
Ghost runs persist when runners die without server notificationcritical
▸
Runners lack heartbeat observability available for workersinfo
▸
Flow run heartbeat events delayed until run completionwarning
▸
HTTP/2 protocol error causes agent to stop after failed retriescritical
▸
Agent uptime decreases progressively before complete failurewarning
▸
Heartbeat fails when Dask worker systemd service lacks virtual environment PATHcritical
▸
Agent heartbeat failure due to docker-in-docker resource exhaustioncritical
▸
Prefect agent freezes in background screen sessioncritical
▸
Flow late-state notifications fail when agent is unresponsivewarning
▸
Agent process alive but unresponsive after prolonged operationcritical
▸
Prefect agents run indefinitely until manually killedinfo
▸
Task heartbeat timeout creates zombie tasks during API outagewarning
▸
Karpenter spot instance termination not detected by Prefect causing stuck flowscritical
▸