feat(cluster): implement SWIM failure detector

Add the core SWIM/Lifeguard failure detection runtime to nodedb-cluster.
The detector drives a tokio::select! probe loop that separates I/O from
logic via a Transport trait, allowing deterministic in-process testing
with InMemoryTransport and edge-drop/partition injection.

- FailureDetector: main runtime task owning the probe loop, suspicion
  timer, and inflight-probe registry
- ProbeRound: Lifeguard ping → ping-req indirect probe sequence with
  per-probe-id oneshot correlation
- ProbeScheduler: random-permutation epoch scheduler (every peer probed
  once per epoch, per Lifeguard §4.3)
- SuspicionTimer: timeout math per Lifeguard §3.1
  (max(min, mult * log2(n) * probe_interval)) with drain_expired polling
- Transport / InMemoryTransport / TransportFabric: Send+Sync async trait
  with mpsc-backed in-memory impl for unit tests
- SwimError variants TransportClosed and ProbeInflightOverflow

async-trait added to nodedb-cluster dependencies to support the
Transport trait definition.

Author: farhan-syah

nodedb-cluster/Cargo.toml

@@ -22,6 +22,7 @@ thiserror = { workspace = true }
 tracing = { workspace = true }
 tokio = { workspace = true }
 rand = { workspace = true }
+async-trait = { workspace = true }
 
 # Serialization (compact binary for RPCs — matches nexar wire format)
 rkyv = { workspace = true }

nodedb-cluster/src/swim/detector/mod.rs

@@ -0,0 +1,20 @@
+//! SWIM failure detector — the runtime that drives the probe loop.
+//!
+//! This module is the `!I/O` portion of the SWIM subsystem: it owns the
+//! probe scheduler, the suspicion timer, and the main `tokio::select!`
+//! loop. All actual networking is pushed behind the [`Transport`] trait
+//! so unit tests can run fully in-process against [`InMemoryTransport`]
+//! and the real UDP transport in E-ε can slot in without touching the
+//! detector logic.
+
+pub mod probe_round;
+pub mod runner;
+pub mod scheduler;
+pub mod suspicion;
+pub mod transport;
+
+pub use probe_round::{ProbeOutcome, ProbeRound};
+pub use runner::FailureDetector;
+pub use scheduler::ProbeScheduler;
+pub use suspicion::SuspicionTimer;
+pub use transport::{InMemoryTransport, Transport, TransportFabric};