Experimental

SFLOW-V5 - sFlow: A Method for Monitoring Traffic in Switched and Routed Networks

Every requirement this repository extracted from SFLOW-V5, the tests bound to it, and what a reader has verified about them. This summary is enrolled and gated by ./le rfc check.

Overview

Positive

what Ze has

Tested both ways6.7%1 of 15 binding obligations

a positive test proves Ze does what the requirement demands and a negative one proves it refuses what the requirement forbids

One polarity plus reason73.3%11 of 15 binding obligations

the requirement admits no counter-case, so one polarity plus a recorded reason is the whole proof available for it

One polarity, unexcused0.0%0 of 15 binding obligations

one direction is tested, the other is neither tested nor excused, and nothing states which

Proven by a recorded break0.0%0 of 13 tagged units

a red was observed once under a recorded procedure, and the unit, the claim and the producer it rested on still hash to what was recorded. The break is not re-run. A test pair is not a proof until one has been observed

Neutral

measures that are neither good news nor bad

Gated MUSTs16of 24 this summary declares

MUST-level requirements the gate HOLDS. A population, not a result: the shares beside it are what says how Ze stands

Out of scope1of 16 gated MUSTs

a {not-applicable} annotation says the obligation does not bind Ze. Scope, not coverage: it is in no share below

Negative

what Ze owes

No test at all20.0%3 of 15 binding obligations

no test carries the requirement id, whether or not a gap states why

The 4 shares marked as a part above are the whole of the 15 obligations that bind Ze: they add to 100%. Proven by a recorded break is a share of TAGGED UNITS, a different population, so it is not one of them.

How to read the colors

A color names what the measure MEANS, not how well Ze scores on it. Green is a good outcome at any value, red is a bad one, and neither a population nor a scope count is an outcome, so both take no color. The number under the label is what says how far Ze has got.

Card	Tone here	Why that color
Gated MUSTs	neutral	no color: a population is a scale, and a larger one is neither good news nor bad. It is the accounting total
Out of scope	neutral	no color: an obligation that never bound Ze is neither an achievement nor a failure, and counting it either way would be a claim
Tested both ways	ok	green at every value: a test pair is the outcome this gate exists to produce, and the share under the label is what says how far Ze has got
One polarity plus reason	ok	green at every value: where no counter-case exists, one polarity IS the complete answer, and a recorded reason is what the gate demands beside it
One polarity, unexcused	ok	green at zero, RED above it: half a proof with no reason for the other half
No test at all	bad	green at zero, RED above it: a binding obligation nothing exercises is a claim with nothing behind it, whether or not a reason is stated
Proven by a recorded break	ok	green at every value: an observed break is the outcome the discrimination gate exists to produce. The denominator is TAGGED UNITS, not obligations, so this share is not one of the parts above
Audit verdicts	warn	RED on the first weak, wrong or unimplemented verdict, amber while a verdict is no longer current or a gated MUST is unjudged, green when every one is judged sound and current

At a glance

Field	Value
Public status	Experimental
Enrolment	Enrolled
Requirements	24
Gated MUST-level	16
Obligations that bind Ze	15
Not applicable, so out of scope	1
Declared gaps	3
Gated with no test	0
Nightly-only evidence	0
Test tags	13
Tagged units	13
Recorded audit verdicts	0
Discrimination records	0
Summary	`rfc/short/sflow-v5.md`
Requirement shard	`rfc/requirements/sflow-v5.md`
RFC text	`rfc/full/sflow-v5.txt`

Enrolment

Enrolled: sFlow Version 5 export: exporter/agent role. 1 MET (counter poll at configured interval) + 11 single-polarity positive (version=5, agent addr, sub-agent seq space, MTU bound, no >1s hold, XDR big-endian, count-prefixed arrays, actual rate, sample_pool, cumulative counters) + 3 gap (split datagram seq, no expanded types, no unavailable sentinel) + 1 not-applicable (collector skip)

What the public ledger says

Status: Experimental

What the ledger says is covered

Flow export protocol alongside NetFlow v9 and IPFIX. Three MUST gaps in rfc/short/sflow-v5.md: datagram-level sequence numbers split across two independent counters per sub-agent (SFLOW-V5-x-9)
no expanded sample types, so ifIndex > 2^24-1 is truncated by the 24-bit source_id mask (SFLOW-V5-x-12)
unavailable if_counters fields are exported as 0 instead of the max-value unavailable sentinel (SFLOW-V5-x-16).

What the ledger says remains:

Coverage

Bucket	Count	What it counts
Positive and negative tests	1	one part of the gated population
Annotated instead of tested	15	one part of the gated population
One polarity only	0	one part of the gated population
No test and no annotation	0	one part of the gated population
Evidence that runs nightly only	0	an overlay: each of these is also counted by the part it falls in
Gated MUST-level requirements	16	every gated MUST falls in exactly one bucket above

Positive and negative tests (1): SFLOW-V5-x-14

Annotated instead of tested (15): SFLOW-V5-x-1, SFLOW-V5-x-2, SFLOW-V5-x-3, SFLOW-V5-x-4, SFLOW-V5-x-5, SFLOW-V5-x-6, SFLOW-V5-x-7, SFLOW-V5-x-8, SFLOW-V5-x-9, SFLOW-V5-x-10, SFLOW-V5-x-11, SFLOW-V5-x-12, SFLOW-V5-x-13, SFLOW-V5-x-15, SFLOW-V5-x-16

Requirements

Requirement	Level	Section	Tests
`SFLOW-V5-x-1`	Datagram version field MUST be set to 5 (Datagram Format)
	MUST	x	positive`unit/verifyTestSFlowDatagramHeaderIPv4` negative`no test`no negative test {single-polarity}: WriteDatagramHeader unconditionally writes the compile-time constant Version=5 into every datagram, so there is no other-version code path to reject (internal/plugins/flowexport/sflow/encoder.go:40, :14)
`SFLOW-V5-x-2`	Agent address MUST be a stable IP (e.g., loopback) that uniquely identifies the device across reboots (Agent Architecture)
	MUST	x	positive`unit/verifyTestSFlowDatagramHeaderIPv4` negative`no test`no negative test {single-polarity}: the operator-configured agent address is written verbatim into every datagram header and validated as a well-formed IP; stability/uniqueness is a config-value property with no exporter reject path (internal/plugins/flowexport/sflow/encoder.go:43-57, internal/plugins/flowexport/config.go:379-383)
`SFLOW-V5-x-3`	Agent address + sub_agent_id MUST uniquely identify a sampling entity (Agent Architecture)
	MUST	x	positive`unit/verifyTestSFlowDatagramHeaderIPv6` negative`no test`no negative test {single-polarity}: both agent_address and sub_agent_id are emitted in every datagram header by construction, and tuple uniqueness is an operator-config obligation (internal/plugins/flowexport/sflow/encoder.go:59, :43-57)
`SFLOW-V5-x-4`	Each sub-agent MUST maintain its own sequence number space (Agent Architecture)
	MUST	x	positive`unit/verifyTestSFlowMultiInterface` negative`no test`no negative test {single-polarity}: each encoder instance is bound to one collector/sub_agent_id and owns private datagramSeq and per-source seqNums fields, so distinct sub-agents never share sequence state (internal/plugins/flowexport/sflow/adapter.go:18-19, flow_adapter.go:32-33)
`SFLOW-V5-x-5`	Datagram size MUST NOT exceed path MTU (Transport)
	MUST	x	positive`unit/verifyTestSFlowMultiInterface` negative`no test`no negative test {single-polarity}: every datagram is bounded to MaxDatagramSize=1400 by construction (counter batching flushes before overflow, the flow encoder truncates the captured header to fit), so no code path emits an oversized datagram (internal/plugins/flowexport/sender.go:14, internal/plugins/flowexport/sflow/encoder.go:120, flow_adapter.go:67-71)
`SFLOW-V5-x-6`	Samples MUST NOT be held more than 1 second before sending (Transport)
	MUST	x	positive`unit/verifyTestExportFlowSampleDispatch` negative`no test`no negative test {single-polarity}: counter and flow samples are encoded and sent synchronously with no buffering queue, so a sample is never held beyond a sub-millisecond encode and there is no holding timer to test negatively (internal/plugins/flowexport/exporter.go:204, internal/plugins/flowexport/sflow/adapter.go:47-51)
`SFLOW-V5-x-7`	All structures MUST use XDR encoding: 4-byte alignment, big-endian (Datagram Format)
	MUST	x	positive`unit/verifyTestSFlowIfCounters` negative`no test`no negative test {single-polarity}: every field is written via binary.BigEndian with 4-byte-aligned opaque padding, exporter-only with no decode path to reject a wrong endianness (internal/plugins/flowexport/sflow/counter.go:36, flow.go:123-128)
`SFLOW-V5-x-8`	Variable-length arrays and opaque data MUST be prefixed by a 4-byte count and padded to 4-byte boundary (XDR Encoding)
	MUST	x	positive`unit/verifyTestSFlowSampledHeader` negative`no test`no negative test {single-polarity}: the sampled_header opaque and the extended_gateway arrays are all written with a 4-byte count prefix and zero-padded to a 4-byte boundary by construction (internal/plugins/flowexport/sflow/flow.go:116-128, :187-208)
`SFLOW-V5-x-9`	Sequence numbers MUST be per-agent (datagram-level) and per-source (sample-level), unsigned 32-bit, wrapping (Transport)
	MUST	x	positive`no test`no positive test negative`no test`no negative test {gap}: per-source sample-level sequences are correct, but for one sub-agent the counter-datagram and flow-datagram streams keep two independent datagramSeq counters, so a collector sees two overlapping datagram-level sequence spaces instead of one (internal/plugins/flowexport/sflow/adapter.go:18, flow_adapter.go:32)
`SFLOW-V5-x-10`	flow_sample MUST include the actual sampling_rate used by the agent (Flow Sample)
	MUST	x	positive`unit/verifyTestSFlowFlowSample` negative`no test`no negative test {single-polarity}: EncodeFlowSample writes the kernel-reported actual rate into every flow_sample, emitted unconditionally with no reject path (internal/plugins/flowexport/sflow/flow_adapter.go:78-79, flow.go:52)
`SFLOW-V5-x-11`	sample_pool MUST track total packets seen by the data source (Flow Sample)
	MUST	x	positive`unit/verifyTestSFlowFlowSamplePoolTracksTotal` negative`no test`no negative test {single-polarity}: EncodeFlowSample computes sample_pool as the saturated product of cumulative samples and rate and writes it into every flow_sample, exporter-only with no negative form (internal/plugins/flowexport/sflow/flow_adapter.go:73-79, flow.go:56)
`SFLOW-V5-x-12`	Expanded sample types MUST be used when ifIndex exceeds 2^24-1 or 2^30-1 (Expanded Flow/Counter Sample)
	MUST	x	positive`no test`no positive test negative`no test`no negative test {gap}: ze implements no flow_sample_expanded (format 3) or counters_sample_expanded (format 4); a large ifIndex is silently truncated by a 24-bit mask on source_id rather than switching to the expanded encoding (internal/plugins/flowexport/sflow/counter.go:49, flow.go:48)
`SFLOW-V5-x-13`	Unknown record formats MUST be skipped using the opaque length prefix (Record Wrappers)
	MUST	x	positive`no test`no positive test negative`no test`no negative test {not-applicable}: skipping unknown formats on receive is a collector behavior; ze is an sFlow exporter only with no sFlow decode path, though it does emit the length prefixes that let a collector skip (internal/plugins/flowexport/sflow/counter.go:60-61, flow.go:131-132)
`SFLOW-V5-x-14`	Counter samples MUST be produced at the configured polling interval for each data source (Counter Polling)
	MUST	x	positive`unit/verifyTestSFlowCounterPollAtInterval` negative`unit/verifyTestSFlowCounterPollBeforeInterval`
`SFLOW-V5-x-15`	All counters MUST be cumulative since boot (Counter Polling)
	MUST	x	positive`unit/verifyTestInterfaceCountersFromCumulative` negative`no test`no negative test {single-polarity}: interfaceCountersFrom copies the raw cumulative kernel counters straight through with no differencing, so exported if_counters are cumulative by construction (internal/plugins/flowexport/register.go:343-358, snapshot.go:10-13)
`SFLOW-V5-x-16`	Unavailable counter fields MUST be set to max value for the type (0xFFFFFFFF for u32, 0xFFFFFFFFFFFFFFFF for u64) (Implementation Guidance)
	MUST	x	positive`no test`no positive test negative`no test`no negative test {gap}: interfaceCountersFrom leaves fields the kernel does not expose (ifInUnknownProtos, ifInBroadcastPkts, ifOutMulticastPkts, ifOutBroadcastPkts) at zero rather than the required max-value unavailable sentinel, so a collector cannot distinguish true-zero from unavailable (internal/plugins/flowexport/register.go:343-361)
`SFLOW-V5-x-17`	Default datagram max size SHOULD be 1400 bytes (Transport)
	SHOULD	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-18`	Default max header size SHOULD be 128 bytes for sampled_header (Raw Packet Header)
	SHOULD	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-19`	Counter polls SHOULD be staggered across sources with randomized initial offset (Counter Polling)
	SHOULD	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-20`	Per-source skip counter SHOULD be initialized to random value in [1, 2*N-1] for 1-in-N sampling (Packet Sampling)
	SHOULD	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-21`	UDP port 6343 SHOULD be used (IANA-assigned) (Transport)
	SHOULD	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-22`	Multiple sFlow instances MAY exist per data source (Agent Architecture)
	MAY	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-23`	Counter samples MAY piggyback on flow sample datagrams when space permits (Counter Polling)
	MAY	x	positive`no test`no positive test negative`no test`no negative test
`SFLOW-V5-x-24`	Agents MAY adjust sampling rates to hardware-supported values; actual rate MUST be reported in each flow sample (Packet Sampling)
	MAY	x	positive`no test`no positive test negative`no test`no negative test

Gaps and untested MUSTs

Requirement	State	Reason
`SFLOW-V5-x-9` Sequence numbers MUST be per-agent (datagram-level) and per-source (sample-level), unsigned 32-bit, wrapping (Transport)	{gap}, no test	per-source sample-level sequences are correct, but for one sub-agent the counter-datagram and flow-datagram streams keep two independent datagramSeq counters, so a collector sees two overlapping datagram-level sequence spaces instead of one (internal/plugins/flowexport/sflow/adapter.go:18, flow_adapter.go:32)
`SFLOW-V5-x-12` Expanded sample types MUST be used when ifIndex exceeds 2^24-1 or 2^30-1 (Expanded Flow/Counter Sample)	{gap}, no test	ze implements no flow_sample_expanded (format 3) or counters_sample_expanded (format 4); a large ifIndex is silently truncated by a 24-bit mask on source_id rather than switching to the expanded encoding (internal/plugins/flowexport/sflow/counter.go:49, flow.go:48)
`SFLOW-V5-x-13` Unknown record formats MUST be skipped using the opaque length prefix (Record Wrappers)	no test	no test carries this requirement id; annotated {not-applicable}: skipping unknown formats on receive is a collector behavior; ze is an sFlow exporter only with no sFlow decode path, though it does emit the length prefixes that let a collector skip (internal/plugins/flowexport/sflow/counter.go:60-61, flow.go:131-132)
`SFLOW-V5-x-16` Unavailable counter fields MUST be set to max value for the type (0xFFFFFFFF for u32, 0xFFFFFFFFFFFFFFFF for u64) (Implementation Guidance)	{gap}, no test	interfaceCountersFrom leaves fields the kernel does not expose (ifInUnknownProtos, ifInBroadcastPkts, ifOutMulticastPkts, ifOutBroadcastPkts) at zero rather than the required max-value unavailable sentinel, so a collector cannot distinguish true-zero from unavailable (internal/plugins/flowexport/register.go:343-361)

Proof state

A tagged unit reads unproven where no discrimination record exists for it: nothing in this tree has been observed to break it, so the claim its tag makes is unproven.