swactor/crates/distribution/tests/telemetry.rs

270 lines
9.6 KiB
Rust
Raw Normal View History

//! Distribution observability contracts.
//!
//! These tests are intentionally narrower than protocol tests: they pin public record/snapshot
//! shapes and assert telemetry remains observational rather than protocol-driving.
//!
//! Behavioral/correctness guarantees:
//! - Distribution telemetry records and snapshot shapes are stable wire contracts.
//! - Telemetry is observational: it reports protocol/cache/registry state without feeding back
//! into behavior.
//! - Transition causes and recent probe targets are externally visible when promised.
mod telemetry_records {
//! Distribution-owned telemetry records keep their channels and JSON payloads stable through
//! the mux.
use distribution::telemetry::{
CacheEntryRec, DIST_STATE, DistributionState, MEMBERSHIP, MembershipTransition,
RegistryEntryRec, SWIM_PROBES, SwimProbeEvent, TRANSPORT_INTERNALS, TransportInternals,
};
use telemetry::frame::{Lifetime, NodeId, StreamId};
use telemetry::{ChannelId, Mux, Position, Record};
#[test]
fn distribution_state_record_round_trips_from_owner_crate() {
let state = DistributionState {
cache_size: 2,
cache_entries: vec![CacheEntryRec {
actor_addr: "actor-a".into(),
node_id: "node-a".into(),
}],
directory_route_count: 7,
registry_size: 3,
registry_tombstones: 1,
registry_entries: vec![RegistryEntryRec {
name: "svc".into(),
actor_addr: "actor-b".into(),
node_id: "node-b".into(),
tombstone: false,
}],
recent_probe_targets: vec!["node-c".into()],
peer_auth_mode: "allow-list".into(),
authorized_peer_count: 4,
};
assert_eq!(DistributionState::CHANNEL, DIST_STATE);
assert_eq!(DistributionState::decode(&state.encode()).unwrap(), state);
}
#[test]
fn distribution_emits_owned_channel_through_telemetry_mux() {
let stream = StreamId::new(NodeId::new("dist-node"), Lifetime(1));
let mux = Mux::unbounded(stream);
let state = DistributionState {
registry_size: 9,
..Default::default()
};
2026-07-18 09:38:16 +00:00
let channel = ChannelId(1);
assert!(mux.submit(channel, state.encode()));
let frames = mux.drain();
assert_eq!(frames.len(), 1);
2026-07-18 09:38:16 +00:00
assert_eq!(frames[0].channel, channel);
assert_eq!(frames[0].position, Position(0));
assert_eq!(
DistributionState::decode(&frames[0].payload)
.unwrap()
.registry_size,
9
);
}
#[test]
fn transport_internals_record_round_trips_probe_rtt() {
let internals = TransportInternals {
relay_connected: true,
direct_peers: 2,
relay_peers: 1,
rtt_ms_p50: 405,
};
assert_eq!(TransportInternals::CHANNEL, TRANSPORT_INTERNALS);
assert_eq!(
TransportInternals::decode(&internals.encode()).unwrap(),
internals
);
}
#[test]
fn membership_transition_record_round_trips_from_owner_crate() {
let transition = MembershipTransition {
peer: "peer-a".into(),
from: "alive".into(),
to: "suspect".into(),
reason: "probe timeout".into(),
last_ack_age_ms: Some(15_000),
consecutive_timeouts: 2,
recent_probe_targets: vec!["peer-a".into(), "peer-b".into()],
member_state: Some("Suspect".into()),
};
assert_eq!(MembershipTransition::CHANNEL, MEMBERSHIP);
assert_eq!(
MembershipTransition::decode(&transition.encode()).unwrap(),
transition
);
}
#[test]
fn swim_probe_event_record_round_trips_from_owner_crate() {
let event = SwimProbeEvent {
event: "timed_out".into(),
target: "peer-a".into(),
sequence: 9,
kind: "direct".into(),
rtt_ms: None,
budget_ms: Some(15_000),
budget_ticks: Some(15_000),
last_ack_age_ms: Some(45_000),
consecutive_timeouts: 3,
recent_probe_targets: vec!["peer-a".into()],
member_state: Some("Suspect".into()),
local_phase: "weights_loaded_wait".into(),
probe_interval_ms: 200,
probe_timeout_ms: 15_000,
indirect_probes: 2,
suspicion_timeout_ms: 45_000,
dead_reprobe_interval_ms: 1_000,
probe_mode: "Periodic".into(),
lifeguard_enabled: false,
};
assert_eq!(SwimProbeEvent::CHANNEL, SWIM_PROBES);
assert_eq!(SwimProbeEvent::decode(&event.encode()).unwrap(), event);
}
}
mod snapshot_and_swim_telemetry {
//! Snapshot and SWIM telemetry readouts: public empty-state shape, recent probe targets, and
//! transition causes.
use distribution::snapshot::DistributionNodeSnapshot;
use distribution::swim::node::{SwimObservation, SwimObserver};
use distribution::swim::telemetry::SwimTelemetry;
use distribution::types::{MemberState, NodeId};
use std::time::Duration;
fn id(byte: u8) -> NodeId {
NodeId([byte; 32])
}
#[test]
fn empty_distribution_snapshot_keeps_the_public_wire_shape_stable() {
// Correctness: external distribution consumers can deserialize the retained
// snapshot shape, and the empty constructor reports explicit zero/open defaults.
let snapshot = DistributionNodeSnapshot::empty(id(0xAB));
let json = serde_json::to_string(&snapshot).unwrap();
let decoded: DistributionNodeSnapshot = serde_json::from_str(&json).unwrap();
assert_eq!(
decoded.node_id,
"abababababababababababababababababababababababababababababababab"
);
assert_eq!(decoded.peer_auth_mode, "open");
assert_eq!(decoded.alive_count, 0);
assert!(decoded.members.is_empty());
assert!(decoded.registry_entries.is_empty());
assert!(decoded.authorized_peer_count.is_none());
}
#[test]
fn swim_telemetry_reports_probe_targets_and_transition_causes_without_state_feedback() {
// Correctness: telemetry is an observer side channel. It records recent probes
// and transition causes for readers, and draining transitions only affects the
// telemetry buffer, not protocol state.
let telemetry = SwimTelemetry::new();
let peer = id(4);
telemetry.observe(SwimObservation::ProbeSent {
target: peer,
sequence: 1,
kind: "direct",
});
telemetry.observe(SwimObservation::Transition {
peer,
from: Some(MemberState::Alive),
to: MemberState::Suspect,
reason: "probe-timeout",
});
assert_eq!(telemetry.rtt_ms_p50(), 0);
assert_eq!(telemetry.recent_targets(), vec![peer]);
assert_eq!(telemetry.last_reasons().get(&peer), Some(&"probe-timeout"));
let drained = telemetry.drain_transitions();
assert_eq!(drained.len(), 1);
assert_eq!(drained[0].peer, peer);
assert_eq!(drained[0].from, Some(MemberState::Alive));
assert_eq!(drained[0].to, MemberState::Suspect);
assert_eq!(drained[0].reason, "probe-timeout");
assert_eq!(drained[0].consecutive_timeouts, 0);
assert!(drained[0].last_ack_age.is_none());
assert!(telemetry.drain_transitions().is_empty());
assert_eq!(telemetry.recent_targets(), vec![peer]);
}
#[test]
fn swim_telemetry_keeps_recent_probe_targets_bounded_and_ordered() {
let telemetry = SwimTelemetry::new();
for byte in 0..17 {
telemetry.observe(SwimObservation::ProbeSent {
target: id(byte),
sequence: byte as u64,
kind: "direct",
});
}
let expected = (1..17).map(id).collect::<Vec<_>>();
assert_eq!(telemetry.recent_targets(), expected);
}
#[test]
fn swim_telemetry_records_probe_events_and_timeout_state_without_fabricating_rtt() {
let telemetry = SwimTelemetry::new();
let peer = id(8);
telemetry.observe(SwimObservation::ProbeSent {
target: peer,
sequence: 1,
kind: "direct",
});
std::thread::sleep(Duration::from_millis(1));
telemetry.observe(SwimObservation::ProbeAcked {
target: peer,
sequence: 1,
kind: "direct",
});
telemetry.observe(SwimObservation::ProbeSent {
target: peer,
sequence: 2,
kind: "direct",
});
telemetry.observe(SwimObservation::ProbeTimedOut {
target: peer,
sequence: 2,
kind: "direct",
budget_ticks: 15_000,
});
let events = telemetry.drain_probe_events();
assert_eq!(events.len(), 1, "only timeout remains immediate");
assert_eq!(events[0].event, "timed_out");
assert_eq!(events[0].budget_ms, Some(15_000));
assert_eq!(events[0].rtt_ms, None);
assert_eq!(events[0].consecutive_timeouts, 1);
assert!(telemetry.drain_probe_events().is_empty());
std::thread::sleep(Duration::from_millis(1_010));
let summary = telemetry
.drain_probe_summary()
.expect("one-second routine probe aggregate");
assert_eq!(summary.sent, 2);
assert_eq!(summary.acked, 1);
assert_eq!(summary.direct_sent, 2);
assert_eq!(summary.rtt_samples, 1);
assert!(summary.rtt_ms_p50.is_some());
}
}