Introduce a read-snapshot plane so pure-snapshot control queries render in
the control-socket task instead of round-tripping the rx_loop, removing the
head-of-line coupling that let a busy or slow rx_loop time out fipsctl and
fipstop observability.
- ControlReadHandle: a cloneable bundle the control accept loop holds, over
the node's already-shared NodeContext and MetricsRegistry plus an
ArcSwap-published StatsSnapshot. A snapshot_dispatch seam serves cut-over
commands off-loop and falls through to the rx_loop for the rest, keeping
the rx_loop's ownership of Node intact.
- StatsSnapshot is published from the tick (the natural and sole mutator of
stats_history), carrying the history rings plus the scalar gauges and
counts show_status reports. Readers serve the latest snapshot
unconditionally, with staleness bounded by the tick interval and no
IO_TIMEOUT-coupled fallback.
- Off-loop now: show_status, show_stats_history, show_stats_all_history,
show_listening_sockets, show_stats_list, and a new counter-only
show_metrics (exposed as fipsctl "stats metrics", the enabler for a
Prometheus scraper at no hot-path cost). Queries that need live per-entity
state (peers, links, sessions, routing, and the per-peer stats variants)
stay on the rx_loop path pending later phases.
Quartet green; forward-merge to next verified clean.
The pool's TTL clock (VirtualIpMapping.last_referenced) advanced only on
DNS re-query, never on traffic, and the mapping-TTL is wired equal to the
DNS TTL, so an in-use mapping was forced to drain at TTL and reclaimed at
the first zero-conntrack tick (a stale drain_start gave no grace effective
protection), breaking long-lived, bursty, or DNS-cached clients.
In tick(), refresh last_referenced whenever conntrack reports sessions > 0
so an actively used mapping never ages out, and recover a Draining mapping
to Active (clearing drain_start) when traffic resumes, so a later drain
gets a fresh grace window instead of a stale one. The Active arm now only
drains an idle mapping. The DNS-TTL / idle-reclaim-TTL wiring is unchanged.
Adds regression tests for continuous-traffic-survives-past-TTL, bursty
drain-then-recover, and fresh-grace-on-redrain.
The per-transport TCP inbound cap was hardwired to 256 and never read
node.limits.max_connections, so raising max_connections was a silent
no-op for inbound TCP. Resolve the effective cap with precedence:
explicit per-transport max_inbound_connections, then node-wide
max_connections, then the built-in default of 256. Established peers
remain bounded node-wide by add_connection, so deriving the per-transport
raw-accept ceiling from max_connections does not admit more real peers
across multiple transports.
Add effective_max_inbound on the TCP transport with a node_max_connections
setter wired from create_transports, plus a precedence unit test.
Estimate the OR-union cardinality over self plus every connected peer's
inbound filter, dropping the parent/child tree gating in
compute_mesh_size. Filter propagation is split-horizon, so cross-links
advertise near-complete mesh views; unioning all peers yields the same
set as the tree-only union in steady state (OR dedups overlap and no
filter can over-count) while damping the node-count flap on parent
switches, since dropping the parent no longer collapses the upward leg.
This also removes the estimate's dependence on tree-declaration cache
freshness.
Rename the debug-log child_count to contributor_count, adapt the two
membership-invariant tests to all-peers semantics, and add a test that
the estimate stays stable across a parent drop when a healthy cross-link
is present.
The inbound FilterAnnounce FPR cap rejects filters whose false-positive
rate (fill^k) exceeds the configured maximum. On the fixed 1 KB / k=5
filter, 0.05 corresponds to fill 0.549 (~1,300 reachable entries), and
the busiest nodes' aggregates were beginning to hit that ceiling as the
mesh grew. Raise the default to 0.10 (fill 0.631, ~1,630 entries) to
restore headroom toward the fixed-filter capacity limit. A saturated or
poisoned filter is ~100% FPR and remains rejected, so the antipoison
gate is not materially weakened.
Updates the config default, the config-reference and bloom-filter
design docs, and the changelog.
A node with a single tree peer has its periodic parent re-evaluation
disabled (it needs at least two peers for a meaningful comparison), so
it depends entirely on its peer pushing a TreeAnnounce for it to attach.
That push happens once at promotion time plus on the parent's slow
periodic no-change re-broadcast. If the one-shot attaching announce is
lost, the single-uplink node falls back to self-root and cannot recover
until the next periodic re-broadcast (reeval_interval_secs later),
stranding it out of the tree and unreachable end-to-end in the interim.
Make tree-position exchange self-healing on the receive path: when an
accepted TreeAnnounce advertises a root strictly worse (higher NodeAddr,
since election is smallest-wins) than our own, echo our current
declaration back to that peer. A stranded self-root node's announce now
provokes its better-rooted peer to re-push its real position immediately,
so the node re-attaches within a round-trip instead of waiting for the
periodic cadence.
Echo only in that one direction. If the peer's root is lower (better)
than ours, we are the stale side: the peer would ignore our worse root
anyway and we converge via the parent re-evaluation that follows, so
echoing back is pure waste and would double announce traffic in the
learning direction during a root change or partition merge. Equal roots
are already converged. The echo is bounded by the existing per-peer
500 ms tree-announce rate limiter and is a no-op once the peer adopts our
root, so it adds no traffic in a converged mesh.
Add a spanning-tree unit test that drives a converged child back to
self-root with its peer-ancestry view of the root cleared (modelling the
lost attaching announce), and asserts the root re-pushes on the
resulting root disagreement and the child re-attaches.
Ignore duplicate or counter-regressed ReceiverReports before updating
RTT, loss, goodput, or ETX, so a delayed or reordered report can no
longer poison link metrics. Compute the RTT-from-echo sample with
checked timestamp arithmetic and reject zero, negative, or out-of-range
results instead of risking wrap or underflow on untrusted wire values.
On the sender side, when receiver dwell time overflows the u16 wire
field, suppress the timestamp echo (send 0) and saturate dwell to
u16::MAX rather than truncating, so a bogus small RTT cannot be formed.
Adds duplicate, out-of-order, wrapped-add, and future-dated (checked_sub)
sample tests, asserts loss and goodput stay unchanged on a dropped
duplicate, and covers the dwell-overflow echo suppression. Documents the
behavior in the MMP design note and CHANGELOG.
Co-authored-by: Johnathan Corgan <johnathan@corganlabs.com>
Brings the maint-line OR-union mesh-size estimator fix and the
per-peer / capacity-cap log-level demotions forward to master.
Conflict resolution: compute_mesh_size resolved to master's config()
accessor form carrying maint's OR-union rewrite (drop the stale summing
`total`); the log-level demotions auto-merged into master's handler
versions. The equivalent master-only log "connected UDP socket installed"
(connected_udp.rs, a file that does not exist on maint) was demoted
info -> debug as part of this merge so the connected-UDP path matches the
rest of the per-peer lifecycle logging.
On a saturated public-mesh node the connection-lifecycle and capacity-cap
events fire continuously and drown out the genuinely notable INFO/WARN
lines. Demote them to debug and drop a redundant duplicate:
- FMP K-bit cutover promotion (encrypted): info -> debug
- "Connection promoted to active peer" (handshake): info -> debug, and
remove the duplicate "Inbound peer promoted to active" line that
shadowed it on the inbound path
- "Peer restart detected" (handshake): info -> debug
- "Peer removed and state cleaned up" (dispatch): info -> debug
- "Rejecting inbound TCP connection (max_inbound_connections reached)"
(tcp): warn -> debug
- "Congestion detected, CE flag set on forwarded packet" (forwarding):
warn -> debug
- "Removing peer: link dead timeout" (mmp): warn -> debug
These are expected, high-frequency conditions on a busy public node (new
and reconnecting peers, ECN CE marking, the inbound connection cap, and
link-dead churn), not operator-actionable signals.
The mesh-size estimator summed the per-filter cardinality of the parent
filter and each child filter, which assumes those filters are perfectly
disjoint. When they overlap -- a stale or oversized parent filter, or a
routing loop -- the sum over-counts and inflates the reported mesh size
to as much as several times the true size.
Estimate the cardinality of the OR-union of the contributing filters
(self + parent + children) once instead. OR is idempotent, so any
overlap is deduplicated: the result equals the old sum in the disjoint
case and stays correct under overlap. The union is seeded from a clone
of a contributing filter so it keeps that filter's size class, and a
filter whose size class does not match is skipped rather than panicking.
The refuse-to-estimate behavior on a saturated or above-cap filter is
preserved.
Add a regression test with overlapping parent and child filters where
the naive sum over-counts and the union estimate tracks the distinct
member count.
The transport layer used Mutex::lock().unwrap() at ten sites across the
UDP, BLE, and Ethernet code. A std mutex poisons if a thread panics
while holding it, after which every lock().unwrap() on that same mutex
also panics, turning one fault into a cascade. These critical sections
only perform short HashMap/Vec operations on locally constructed values
and are not reachable from peer input, but the idiom is fragile against
any future in-section panic. Replace each with
lock().unwrap_or_else(|e| e.into_inner()), which recovers the guarded
data and removes the cascade with no new dependency and no call-graph
change.
Also replace four self.local_addr.unwrap() calls in the UDP start and
adopt paths with a sentinel fallback. The value is provably set just
above each log line today, but the unwrap is brittle against a future
reordering; logging an unbound sentinel is harmless and cannot panic.
When two peers each auto_connect to the other, each runs both an
initiator and a responder NAT-traversal session and binds a separate
UDP socket per session. Each side adopts only the first Established
event and drops the loser session's socket; when the two sides adopt
mismatched sessions, each sends its Noise msg1 to a peer port the peer
has already stopped draining, and both handshakes stall.
Deterministically keep the session initiated by the smaller NodeAddr,
decided on the responder path: decline an incoming offer only when we
also have an in-flight outbound initiator for the same peer and our
NodeAddr is smaller. The peer's redundant initiator then times out,
leaving a single matching socket pair on both ends. Asymmetric
(one-sided) auto_connect has no co-active initiator and is never
suppressed, so connectivity is preserved; an undecidable NodeAddr falls
through to answering.
Reuses the NodeAddr tie-breaker convention already used by the
cross-connection and rekey dual-init paths. Adds a unit test for the
election helper.
Add a Loopback variant to TransportHandle backed by an unbounded
in-process channel and a shared address-to-receiver registry, so
node-level multi-node tests deliver packets directly between nodes
instead of over real localhost UDP sockets. This removes the kernel
UDP receive-buffer overflow that dropped handshake packets when many
tests ran in parallel under CPU contention, and lets the large-network
convergence tests run reliably in the default suite again (their
parallel-load ignore markers are removed).
The new transport and its enum variant are cfg(test)-gated, so the
daemon build is unaffected.
The FMP rekey msg1 resend driver retransmitted indefinitely with no cap
and no abandon, so a rekey that never completed kept resending msg1
forever. Give it a retransmission budget: cap resends at
handshake_max_resends with exponential backoff and abandon the rekey
cycle cleanly once the budget is exhausted, mirroring the FSP session
rekey msg3 driver.
With the cap in place the link-dead heartbeat can safely become
rekey-aware: check_link_heartbeats now suppresses teardown while a rekey
is in progress with msg1 budget remaining, instead of reaping a link
that is still actively carrying rekey-handshake traffic. The suppression
terminates deterministically (the budget abandons on exhaustion, cutover
clears the in-progress flag), so a genuinely dead link is still reaped on
the next cycle.
Adds a rekey_msg1_resend_count counter on ActivePeer reset at every
rekey-clear and cutover site, msg1 resend-budget unit tests, and two-node
heartbeat suppression/resume/regression integration tests.
Remove the duplicated immutable fields (config, identity, startup_epoch,
started_at, is_leaf_only, max_connections/peers/links) from the Node
struct so the Arc<NodeContext> bundle is the single source of truth.
Previously Node owned these fields and a parallel context copy, kept in
lockstep by rebuild_context() at every mutation site — pure overhead that
existed only because of the duplication.
- Replace rebuild_context() with replace_context(): a clone-edit-swap of
the whole Arc. The per-instance context stays immutable; mutation swaps
the Arc. This is the sole runtime mutation path (constructors, leaf_only,
update_peers).
- Add Copy-returning accessors startup_epoch() and max_connections()/
max_peers()/max_links(); migrate the remaining direct field readers onto
the accessors. node_addr()/npub()/Debug now read identity/is_leaf_only
from the context.
- update_peers reads the pre-update peer set from the live context Arc
before building a fresh Config + context and swapping — preserving the
read-before-write ordering its mutation-window test depends on.
- Remove the test-only set_max_* setters; tests set the limits on Config at
construction instead (new make_node_with_max_peers/links helpers).
- Add a ci-local guard that fails if the Node struct re-declares a bundled
field, so the single-store invariant can't silently regress.
cargo test --lib 1291/0; clippy -D warnings and release build clean.
Store node counters in an atomic metric registry read through &self, and
introduce a shared NodeContext bundle holding the effectively-immutable
fields (config, identity, startup epoch, capability limits). Source the
immutable config and identity reads across the receive hot path, the
handshake/session/mmp/encrypted state machines, and the discovery, tree,
bloom, retry, and lifecycle modules through the context accessors rather
than direct field reads. The Node fields and the context are rebuilt in
lockstep at every mutation site.
PeerRecvDrain::drop previously called std::thread::join on the worker
thread synchronously. The worker uses packet_tx.blocking_send on a
tokio mpsc Sender, which internally parks the worker via
tokio::block_on on the same current_thread runtime that drives
rx_loop. Calling join from inside remove_active_peer (which runs on
the runtime thread, the runtime's sole driver) created a circular
wait:
- rx_loop blocks in libc futex via Thread::join
- the worker being joined cannot observe the stop flag because the
runtime that polls it is the very thread now blocked joining it
- all other PeerRecvDrain workers park on the same runtime via
block_on, so a single peer's removal wedges every worker on the
daemon
The /proc snapshot from a production wedge showed exactly this
shape: 107 of 108 threads in futex_do_wait, 101 of them named
fips-peer-drain. fipsctl became unresponsive (EAGAIN on control
socket), SIGTERM was ignored, and Docker SIGKILLed the container
after the 10 s grace period. Two confirmed wedges on the public
test deployment (52 min and 23 min uptime), plus a third on the
admission-gate-Msg2-silent-drop build at 2 min 21 sec — all ending
with the identical "Peer removed and state cleaned up
tree_changed=false" final log line preceding total silence.
Fix: detach the std::thread instead of joining. The stop flag plus
self-pipe write already signal the worker to exit; the worker's
kernel-level libc::poll inside the drain loop sees the wake, checks
the flag, exits, and the OS reclaims the thread state independently
of the JoinHandle being dropped.
The trigger was statistically amplified by aggressive multi-npub-
from-one-NAT peer reconnect patterns at the moment of the 30 s
link-dead-timeout peer-removal, but not bounded to them. Any
peer whose disconnect happens with the per-peer drain worker
parked in block_on can fire the bug. The admission-gate work
that landed earlier in this branch line compressed more handshake
work per rx_loop tick, increasing the rate at which workers are
parked in block_on and so reducing time-to-wedge — but the
underlying bug pre-dated the admission gate and pre-dated this
fix branch.
The deployed wedged daemon is mitigated operationally by blocking
the trigger IP at the host firewall; this commit removes the bug
class entirely.
The discovery request dedup cache (recent_requests) silently dropped
LookupRequests once it reached MAX_RECENT_DISCOVERY_REQUESTS, with no
counter to surface the condition. Add a DiscoveryReject::ReqDedupCacheFull
reject reason backed by a req_dedup_cache_full counter on DiscoveryStats,
mirroring the existing duplicate-request counter, and record it at the
drop site so the rejection is visible in show_routing.
Bring the runtime peer-list refresh and opt-in mDNS LAN discovery work
on master into the receive-path RejectReason / reloadable-config
integration branch. Code files auto-merge clean; the only conflict is
the CHANGELOG Unreleased section, resolved as the union of both sets of
entries.
Add scoped mDNS / DNS-SD discovery for peers on the same local link,
giving sub-second pairing without a relay or NAT-traversal roundtrip.
A node advertises its npub, protocol version, and an optional network
scope over link-local multicast, and browses for matching adverts to
initiate Noise handshakes against same-LAN peers.
LAN discovery is disabled by default; operators enable it with
node.discovery.lan.enabled: true. Default-off avoids reintroducing a
per-LAN identity broadcast on nodes that have deliberately disabled
other discovery channels, and avoids any multicast surprise on upgrade.
The startup advertised-port picker now excludes bootstrap transports
and selects a non-bootstrap operational UDP transport with a stable
lowest-id selector, so the advertised port is deterministic across
restarts rather than dependent on HashMap iteration order. This
matches the per-dial transport selection used for discovered peers.
Co-authored-by: Johnathan Corgan <johnathan@corganlabs.com>
Add Node::update_peers for runtime peer-list refresh. It re-derives the
active peer connections from a new peer configuration, adding newly
configured peers and removing those no longer present, while keeping
links to peers that remain in the set rather than tearing every
connection down. The call returns an UpdatePeersOutcome summarizing the
added, removed, and retained peers.
PeerAddress gains a seen_at_ms recency field (with_seen_at_ms). Active
path selection now sorts address candidates by recency so the most
recently observed address wins when concurrent path probes race.
complete_rekey_msg2 now returns the remote peer's startup epoch
alongside the new Noise session, letting the rekey path detect a peer
restart and clear stale session state. A stale FSP session is cleared
when a peer restart is detected during FMP rekey or cross-connection
promotion, so the session-layer map no longer lingers out of sync with
the freshly promoted peer.
Per-tick work budgets bound the connection churn in a single node tick
(MAX_DISCOVERY_CONNECTS_PER_TICK, MAX_RETRY_CONNECTIONS_PER_TICK,
MAX_PARALLEL_PATH_CANDIDATES_PER_PEER); work beyond a tick's budget is
deferred to the next tick rather than discarded.
Co-authored-by: Johnathan Corgan <johnathan@corganlabs.com>
Move PeerAclReloader onto the Reloadable trait: its ACL snapshot is now
published through an arc_swap::ArcSwap so the authorization hot path reads
it without locking, and the former check_reload becomes the trait's
reload(). The node tick calls self.peer_acl.reload().await.
Wire the host map into the tick as well. The host map snapshot was
previously taken once at construction and never polled; it now hot-reloads
on /etc/fips/hosts mtime changes once per tick, alongside the ACL, so
hostname display reflects edits without a restart.
The path_mtu_lookup cache (event-driven, populated from observed traffic)
and the nostr_discovery subsystem (an async spawned task) are deliberately
left off the trait: neither reloads from a backing file, so a no-op reload()
would be misleading. The rationale is documented on the trait module.
The host map and the ACL's embedded alias reloader still stat /etc/fips/hosts
independently each tick. A single small-file stat per tick is cheap, so the
duplicate is left in place; sharing one mtime observation between the two is
a possible future cleanup.
Tests: a node-level test exercises the host-map tick reload end to end
through peer_display_name; the ACL reloader tests are updated to drive the
async reload().
Add a `Reloadable` trait that normalizes the node's reloadable
configuration/resource pattern onto a single contract built around an
`arc_swap::ArcSwap` snapshot: a lock-free `load()` for the hot read path
and an async `reload()` that re-reads the backing source and atomically
swaps in a fresh snapshot. The trait carries the canonical Arc-wrapper
template documentation (single-writer node tick, many-reader hot path,
whole-snapshot swap so readers never observe a partial update).
Migrate the host map to this trait via a new `HostMapReloadable` that
reuses the existing load/merge/mtime helpers in upper::hosts. The Node
`host_map` field changes from `Arc<HostMap>` to `HostMapReloadable`, and
`peer_display_name` reads through a lock-free guard. The initial snapshot
is byte-identical to the previous construction, so behavior is unchanged.
The host map is still snapshotted once at construction and not polled;
`reload()` is exercised only by unit tests for now. Wiring the periodic
poll into the node tick, and deduplicating the hosts-file stat against
the ACL reloader's embedded copy, is left as a follow-up.
Add `arc-swap` as a dependency. Unit tests cover initial load (base +
file, base only), change/no-change/deletion/creation detection,
base-preserved-on-reload, and equivalence of the initial snapshot to the
pre-migration construction.
Introduce a typed RejectReason enum and a NodeStats::record_reject
dispatch so every receive-path rejection-and-return site bumps a
machine-readable per-subsystem counter while keeping its operator-facing
log line. The top-level variants mirror the existing NodeStats subsystem
split (Tree, Bloom, Discovery, Forwarding) and add Handshake, Session,
Mmp, and Transport categories; HandshakeStats, SessionStats, and MmpStats
are new sub-stats.
Wired clusters: tree and MMP outbound sign-failure; the FSP session
unknown-session and state-machine cluster; the Noise IK handshake
state-machine cluster (msg1/msg2); and the decode / crypto / cap /
semantic tail across bloom, discovery, forwarding, mmp, and tree. The
TreeStats::ancestry_invalid counter, present since the scaffold but never
incremented, is now bumped from the validate_semantics ancestry rejection.
Several handshake, MMP, tree, and discovery paths that previously had no
counter at all are now counted, including the send_lookup_response
no-route drop (DiscoveryStats::resp_no_route).
Existing direct counters at the bloom / discovery / forwarding sites are
retained alongside the new dispatch while the rollout is in progress (the
bloom_poison tests expect the transitional +2 delta); a later change
collapses the duplicate increment.
Raise the in-process backpressure headroom in make_test_node_with_mtu
(request an 8 MiB recv_buf_size on UdpConfig and grow packet_channel from
256 to 8192) to reduce localhost-UDP receive overflow under parallel-CPU
scheduler contention, and mark the large-network convergence tests
#[ignore] so cargo test --lib stays green by default. The ignored tests
remain runnable on demand with --ignored or --test-threads=1.
Add per-direction pool_inbound/pool_outbound counters to TcpStats and
TorStats, updated at every pool-insert, receive-loop-exit, transport-stop,
and send-failure removal site. Compare the max_inbound_connections cap
against pool_inbound rather than the combined pool length, so outbound
connect-on-send connections no longer consume the operator-facing inbound
budget. The configuration field name and operator semantics are preserved;
only the cap-check comparison and accounting change.
Move the max_peers cap check in handle_msg1 forward, from the late
check inside promote_connection (which fires after Msg2 has already
been built and put on the wire) to an early position after identity
verification but before index allocation and the Msg2 send. When the
gate fires for a net-new identity, the Msg1 is silent-dropped — no
response goes back to the peer, no AEAD compute or wire bytes are
spent.
Bypass preserved for known peers (reconnect / cross-connection): if
the sender's NodeAddr is already in self.peers, or if a pending
outbound connection is in flight to the same identity, the gate is
skipped so legitimate maintenance traffic continues to work. The
late check inside promote_connection is intentionally retained as
defense-in-depth against future call sites or a disconnect racing
between the early-gate decision and promotion.
Wire-cost rationale: a 45 s tcpdump at saturation observed ~3.6
cap-denials/s steady-state, each previously paying the full Noise IK
responder crypto + Msg2 (~104 B) on the wire before being rejected.
The bigger value is cleaner peer-side semantics — the peer no longer
sees a fake-completed handshake whose data frames subsequently fail
decryption locally.
Two new unit tests cover the cases:
- handle_msg1_silent_drops_at_cap_for_new_peer drives a wire-pumped
Msg1 from a fresh identity into a saturated node and asserts no
Msg2 reaches the sender socket. Stash-verifies as FAIL on the
pre-fix tree (Msg2 hits the wire) and PASS post-fix.
- handle_msg1_admits_existing_peer_at_cap drives a Msg1 from an
identity already in self.peers and asserts the gate does not evict
it. This is a regression check (the no-gate tree behaves the same
way here, but the test guards against an accidental future gate
that breaks known-peer admit).
Brings two structural fixes landed on maint:
- compute_mesh_size: explicit parent skip in the children loop, so the
disjoint-subtree invariant no longer depends on peer_declaration cache
freshness.
- max_peers: outbound connection-initiation gated on the cap (auto-reconnect
retries, Nostr-mediated discovery established adoption, and both sides
of the NAT-traversal punch sequence). Inbound msg1 admission gate
unchanged.
node.limits.max_peers was honored only on inbound msg1 admission
(handshake.rs handle_msg1 returns PeerLimitExceeded when peers.len
is at the cap). Four outbound initiation paths proceeded unconditionally
at capacity: auto-reconnect retries (process_pending_retries),
Nostr-mediated discovery's BootstrapEvent::Established adoption
(poll_nostr_discovery), NAT-traversal punch initiation (the outgoing
side of the offer/answer/punch sequence in the Nostr discovery
runtime), and NAT-traversal punch response (the incoming side of the
same sequence). A saturated node burned CPU, UDP probes, STUN
observations, and Nostr relay traffic on connections that the inbound
gate would reject the moment they reached msg1.
Introduce Node::outbound_admission_check (peers.len < max_peers, or
true when max_peers == 0 as the no-cap sentinel) and gate the four
paths. The discovery runtime lives in a separate task and does not
hold a Node reference; bridge via an Arc<AtomicBool> the runtime
reads and Node refreshes once per tick from outbound_admission_check.
The atomic granularity is intentionally loose: one-tick lag is
acceptable because the inbound msg1 gate continues to be the
authoritative cap, and in-flight handshakes started below the cap
are allowed to complete.
Inbound gate at handshake.rs is unchanged.
The mesh-size estimator's children loop relied on the cached
peer_declaration(parent_id).parent_id() != my_addr check to exclude
the parent. That cached view briefly disagrees with our own latest
my_declaration().parent_id() during the window between a local
parent-switch and the new parent's next inbound TreeAnnounce: the
peer-declaration cache still names us as the parent's parent, so the
parent is iterated as if it were a child and its (typically dominant)
bloom cardinality is added a second time. Symptom: estimated mesh size
displayed in fipsctl show status and fipstop nearly-but-not-exactly
doubles during tree rebalancing.
Make the invariant structural with an explicit peer_addr == parent_id
skip at the head of the children loop. Per-peer 500 ms rate-limiter
and overall recompute cadence are unchanged.
Adds a regression test that constructs the stale-peer-declaration
scenario directly and asserts the parent is not double-counted.
Closes the eventually-consistent gap in spanning-tree state
distribution. Every existing send_tree_announce_to_all call site
gates on a local state-change event (parent switch, self-root
promotion, ancestry change, peer promotion, parent loss). Once a
partition latches — for example a parent-switch announce stranded
in the brief cross-init handshake swap window, where the announce
arrives on a session-index whose decrypt-worker entry has been
unregistered — neither side's state changes again, so neither
side ever re-broadcasts. The existing 60 s check_periodic_parent_reeval
was a re-evaluation, not a re-broadcast: it short-circuited
silently on no-change. Production-side healing depended on
incidental link churn; lab harnesses with stable docker-bridge
links had no equivalent path.
Add a final else branch that fires send_tree_announce_to_all
unconditionally on the no-change path, alongside the existing
switch and self-promote arms. Receivers coalesce by sequence
comparison (ParentDeclaration::is_fresher_than) and short-circuit
at the `if !updated` gate in handle_tree_announce; same-sequence
repeats drop silently with no cascade. The per-peer 500 ms
rate-limiter is well below this 60 s cadence and does not suppress
the heartbeat broadcast.
The fix is a general protocol-robustness improvement: it addresses
any in-flight TreeAnnounce loss class, not only the specific
cross-init swap-window drop site.
testing/static/scripts/rekey-test.sh BASELINE_CONVERGENCE_TIMEOUT
60 -> 65 so a partition healed by the periodic broadcast at T+60
lands inside the convergence window. wait_for_full_baseline
early-exits on PASS, so successful reps see no extra wall-clock.
The cross-connection-won path in handle_msg1 removes the old peer and frees
its allocated index, but does not unregister the old (transport_id, our_index)
cache_key from the decrypt worker pool. The orphan entry sits in the
per-shard HashMap until the index allocator recycles old_idx to a different
peer and that peer's register_decrypt_worker_session call overwrites it.
In the interim, any decrypt job that lands at the recycled cache_key
resolves to the wrong session and AEAD silently fails — observed as
multi-hop routing failure in 5-node static-mesh on next-branch where
bidirectional auto_connect drives cross-connections at every peer pair
on startup.
The tick body's per-peer check_* loops (heartbeats, bloom
announces, MMP reports, tree announces) called transport.send
for every active peer, which on TCP/Tor fell through to a 5 s
connect-on-send wait for any peer whose pool entry was not yet
established. That wedged the entire tick body for the full
connect_timeout_ms per unreachable peer; under post-restart
convergence on a high-peer mesh, this cascaded into multi-
second tick stalls. On master, the same mechanism also starved
the per-tick control-snapshot republish and pushed fipsctl
queries onto an mpsc fallback that was itself queued behind
the wedged rx_loop, producing the 5-second fipsctl head-of-line
pattern operators observed on loaded nodes.
Gate send_encrypted_link_message_with_ce on
transport.connection_state before the send: proceed only when
Connected; on None, kick off a non-blocking background connect
(idempotent — TransportHandle::connect dedupes against the
connecting pool and spawns the timeout-bounded TcpStream::connect
inside its own tokio task) and fail this send fast with a
clear "transport connection not ready" error. A subsequent
tick retries once the pool has an entry. The reconnect
lifecycle (check_link_heartbeats, process_pending_retries,
poll_pending_connects) is unchanged. The connect-on-send
branch in transport.send_async itself remains in place for
code paths that legitimately need synchronous connect (e.g.,
explicit operator-driven fipsctl connect).
Johnathan Corgan
Martti MalmiandJohnathan Corgan
