Enum DaemonEvent 

#[non_exhaustive]
pub enum DaemonEvent {
    BuildResult {
        build_id: BuildId,
        result: Result<(), String>,
    },
    SpawnResult {
        dataflow_id: DataflowId,
        result: Result<(), String>,
    },
    AllNodesReady {
        dataflow_id: DataflowId,
        exited_before_subscribe: Vec<NodeId>,
    },
    AllNodesFinished {
        dataflow_id: DataflowId,
        result: DataflowDaemonResult,
    },
    Heartbeat {
        ft_stats: Option<FaultToleranceSnapshot>,
    },
    ZenohListenEndpoint {
        endpoint: Option<String>,
    },
    StatusReport {
        running_dataflows: Vec<DataflowStatusEntry>,
    },
    Log(LogMessage),
    Exit,
    NodeMetrics {
        dataflow_id: DataflowId,
        metrics: BTreeMap<NodeId, NodeMetrics>,
        network: Option<NetworkMetrics>,
    },
    TopicDebugData {
        dataflow_id: DataflowId,
        subscription_ids: Vec<Uuid>,
        payload: Vec<u8>,
    },
    StateCatchUpAck {
        dataflow_id: DataflowId,
        ack_sequence: u64,
    },
    NodeStopped {
        dataflow_id: DataflowId,
        node_id: NodeId,
        clean_stop: bool,
    },
}

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

BuildResult

Fields

build_id: BuildId

result: Result<(), String>

SpawnResult

Fields

dataflow_id: DataflowId

result: Result<(), String>

AllNodesReady

Fields

dataflow_id: DataflowId

exited_before_subscribe: Vec<NodeId>

AllNodesFinished

Fields

dataflow_id: DataflowId

result: DataflowDaemonResult

Heartbeat

Fields

ft_stats: Option<FaultToleranceSnapshot>

ZenohListenEndpoint

The zenoh endpoint this daemon actually bound and is reachable at, sent once its session is open.

The coordinator records it and hands it to daemons that register later (see RegisterResult::Ok::peer_zenoh_endpoints), which is what lets a multi-machine deployment wire itself without every daemon being told every other daemon’s address.

Confirms or withdraws the endpoint this daemon advertised in its registration, once its zenoh session is open and the listener has been verified against info().locators().

Some(endpoint) confirms (and would correct a differing one); None withdraws, which is what a daemon whose listener did not bind must do so the coordinator stops handing out a dead endpoint.

The registration carries the endpoint in the first place — see DaemonRegisterRequest::zenoh_listen_endpoint for why it cannot wait until here. This is the correction, not the announcement.

Fields

endpoint: Option<String>

StatusReport

Sent by the daemon after registration to report its current state. Enables coordinator-daemon reconciliation on reconnect.

Fields

running_dataflows: Vec<DataflowStatusEntry>

Log(LogMessage)

Exit

NodeMetrics

Fields

dataflow_id: DataflowId

metrics: BTreeMap<NodeId, NodeMetrics>

network: Option<NetworkMetrics>

TopicDebugData

Topic debug payload destined for one or more active CLI subscriptions.

Daemon and coordinator are co-deployed from the same build, so this multi-subscriber shape is safe to evolve within the repository.

Fields

dataflow_id: DataflowId

subscription_ids: Vec<Uuid>

payload: Vec<u8>

StateCatchUpAck

Daemon acknowledges state catch-up through a given sequence number.

Fields

dataflow_id: DataflowId

ack_sequence: u64

NodeStopped

Sent by the daemon when a node has exited and the daemon will NOT restart it (e.g. dora node stop, a node exiting under restart_policy: Never, or a final-failure cascade). The coordinator uses this to invalidate its cached node_metrics entry so dora node list reflects the actual state instead of the last-reported “Running” snapshot. Without this signal the daemon’s metrics-snapshot loop simply stops including the dead node and the coordinator’s cache is frozen at the last pre-exit values forever.

Fields

dataflow_id: DataflowId

node_id: NodeId

clean_stop: bool

true if the daemon called disable_restart() before the exit (i.e. the stop_single_node / restart_single_node path triggered by dora node stop/restart). false for a final-failure exit under restart_policy: Never or a max_restarts exhaustion. The coordinator uses this to pick NodeStatus::Stopped vs NodeStatus::Failed, so a crash is not silently reported as a clean teardown (which would hide it from dora doctor).

Trait Implementations

impl Debug for DaemonEvent

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for DaemonEvent

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for DaemonEvent

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.