mirror of https://github.com/jmcorgan/fips.git synced 2026-07-22 07:48:26 +00:00

Files

Martti Malmi d548add18d Reject stale MMP receiver reports

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>

2026-06-07 23:30:35 +00:00

9.5 KiB

Raw Permalink Blame History

Metrics Measurement Protocol (MMP)

The Metrics Measurement Protocol provides per-link and per-session quality metrics — SRTT, loss, jitter, goodput, ETX, and one-way delay trend — using only counter and timestamp fields already present in the FMP and FSP wire formats. No additional probing traffic is required. The same algorithms and report message format are used at both layers; only the routing scope and configuration namespace differ.

This document is the canonical home for the MMP design. For the link-layer instance's role inside FMP, see fips-mesh-layer.md. For the session-layer instance's role inside FSP, see fips-session-layer.md. For the byte-level SenderReport and ReceiverReport layouts, see ../reference/wire-formats.md.

Two Layers, One Protocol

MMP runs at two layers:

Link-layer MMP: One instance per active FMP peer link. Reports are exchanged peer-to-peer between direct neighbors and measure the quality of that single hop.
Session-layer MMP: One instance per established FSP session. Reports are encrypted end-to-end and forwarded through every transit link, measuring end-to-end quality independent of hop count.

The algorithms (SRTT estimation, jitter computation, loss inference, ETX) are identical at both layers. The differences are configuration namespace, report intervals, and routing scope. See Layer Differences below.

Metrics Tracked

MMP computes the following metrics from the per-frame counter and timestamp fields:

SRTT — Smoothed round-trip time (Jacobson/RFC 6298, α=1/8). Derived from timestamp-echo in ReceiverReports with dwell-time compensation.
Loss rate — Bidirectional loss inferred from counter gaps. Tracked as both instantaneous (per-interval) and long-term EWMA.
Jitter — Interarrival jitter (RFC 3550 algorithm) in microseconds.
Goodput — Bytes per second of payload data (excludes MMP reports).
OWD trend — One-way delay trend (µs/s, signed). Indicates congestion buildup before loss occurs.
ETX — Expected Transmission Count, computed from bidirectional delivery ratios. Used in cost-based parent selection via link_cost = etx * (1.0 + srtt_ms / 100.0), and in bloom-filter candidate ranking inside find_next_hop() (the same link_cost is the primary key when choosing among bloom-filter peers, with tree distance as the tie-breaker).
Dual EWMA trends — Short-term (α=1/4) and long-term (α=1/32) trend indicators for both RTT and loss, enabling change detection.

Session-layer MMP additionally tracks the observed forward-path MTU; see fips-mtu.md for the end-to-end path-MTU mechanism.

Operating Modes

MMP supports three modes:

Mode	Reports Exchanged	Metrics Available
Full (default)	SenderReport + ReceiverReport	All metrics including RTT, loss, jitter, goodput, OWD trend
Lightweight	ReceiverReport only	Loss (from counter gaps), jitter, OWD trend. No RTT.
Minimal	None	Spin bit and CE echo flags only. No computed metrics.

The mode is configured per layer (node.mmp.mode and node.session_mmp.mode).

Report Scheduling

Reports are sent at RTT-adaptive intervals computed as clamp(2 × SRTT, low, high). A cold-start interval is used until SRTT has converged.

Layer	Adaptive bounds	Cold-start
Link	`[1s, 5s]`	200 ms (first 5 samples)
Session	`[500ms, 10s]`	1 s

The session-layer bounds are higher because session reports are encrypted and forwarded through every transit link, so bandwidth cost is proportional to path length.

Spin Bit and RTT

The SP (spin bit) flag in the FMP inner header follows the QUIC spin bit pattern: reflected on receive, toggled on send when the reflected value matches the last sent value. The spin bit state machine runs for TX reflection, but RTT samples from the spin bit are discarded. In a mesh protocol where frames are sent irregularly (tree announces, bloom filters, MMP reports on different timers), inter-frame processing delays inflate spin bit RTT measurements unpredictably. Timestamp-echo from ReceiverReports (with dwell-time compensation) is the sole SRTT source.

Duplicate or regressed ReceiverReports are ignored before any RTT, loss, goodput, or ETX update. If receiver-side dwell time exceeds the wire field, the report keeps its counters but sends a zero timestamp echo so the sender cannot form an invalid RTT sample.

The spin bit lives in the link-layer FMP inner header, so this mechanism applies to link-layer MMP only. Session-layer MMP carries its spin bit in the FSP encrypted inner header but uses it the same way: reflected for diagnostic visibility, not used for SRTT.

ECN Congestion Signaling

The CE (Congestion Experienced) flag (bit 1 in the FMP flags byte) provides hop-by-hop congestion signaling through the mesh. Transit nodes detect congestion on outgoing links and set CE on forwarded packets; once set, the flag stays set for all subsequent hops to the destination.

Congestion detection triggers on any of:

Outgoing link MMP loss rate ≥ node.ecn.loss_threshold (default 5%)
Outgoing link MMP ETX ≥ node.ecn.etx_threshold (default 3.0)
Kernel receive buffer drops detected on any local transport (via SO_RXQ_OVFL on UDP)

CE relay: The forwarding path computes outgoing_ce = incoming_ce || local_congestion. Once CE is set on a packet, it remains set for the rest of the forward path.

IPv6 ECN-CE marking: When a CE-flagged DataPacket arrives at its final destination, the IPv6 Traffic Class ECN bits are marked CE (0b11) before TUN delivery — but only for ECN-capable packets (ECT(0) or ECT(1)). Not-ECT packets are never marked per RFC 3168. The host TCP stack then echoes ECE in ACKs, triggering sender cwnd reduction through standard congestion control.

Session-layer tracking: The ecn_ce_count field in MMP ReceiverReports tracks CE-flagged packets received per link, providing end-to-end visibility into congestion propagation.

ECN signaling is a link-layer mechanism. Session-layer MMP only observes the CE counter as part of the report stream; CE marking is not generated end-to-end. Tuning parameters live under node.ecn.* in ../reference/configuration.md.

Send Failure Backoff (Session Layer Only)

When a session MMP report cannot be delivered (destination unreachable, no route), the sender applies exponential backoff to the probe interval — a standard distributed-systems pattern for transient failure handling:

Each consecutive failure doubles the interval: 2x, 4x, 8x, 16x, 32x
Backoff caps at 32x the base interval (5 consecutive failures)
A successful send resets to the normal SRTT-based interval
Debug logging is suppressed after 3 consecutive failures; a summary is logged when the destination becomes reachable again

This prevents wasted CPU and log noise when a session's remote endpoint has departed the network but the local session has not yet timed out. Link-layer MMP has no equivalent — link-layer reports are peer-to-peer over an authenticated link, so delivery failure is indistinguishable from link death and the link-liveness mechanism takes over.

Layer Differences

Aspect	Link layer	Session layer
Routing scope	Peer-to-peer (one hop)	End-to-end (forwarded through every hop)
Configuration namespace	`node.mmp.*`	`node.session_mmp.*`
Report bounds	`[1s, 5s]`	`[500ms, 10s]`
Cold-start interval	200 ms (first 5 samples)	1 s
Bandwidth cost	One link	Proportional to path length
Send-failure backoff	Not applicable	Yes
Path-MTU echo	Not applicable	PathMtuNotification (see fips-mtu.md)
Idle-timeout interaction	None	Reports do not reset session idle timer

Idle Timeout Interaction (Session Layer Only)

MMP reports (SenderReport, ReceiverReport) and PathMtuNotification do not reset the session idle timer. Only application data (DataPacket, type 0x10) resets last_activity. This ensures sessions with no application traffic tear down after node.session.idle_timeout_secs (default 90s), while MMP continues providing measurement data up to the teardown moment.

Operator Logging

Both layers emit periodic metrics at info level. The interval is node.mmp.log_interval_secs for link-layer (default 30s) and node.session_mmp.log_interval_secs for session-layer (default 30s).

Link-layer:

MMP link metrics peer=node-b rtt=2.3ms loss=0.2% jitter=0.1ms goodput=76.0MB/s tx_pkts=1234 rx_pkts=5678

Session-layer:

MMP session metrics session=npub1tdwa...84le rtt=4.3ms loss=0.6% jitter=0.2ms goodput=71.3MB/s mtu=1472 tx_pkts=1234 rx_pkts=5678

Teardown logs include final SRTT, loss rate, jitter, ETX, goodput, and cumulative tx/rx packet and byte counts.

9.5 KiB Raw Permalink Blame History Unescape Escape