The rename check rewrites a bare pre-rename binary name in any processed
doc, which turned the sampler's selector description into "against xrpld
or xrpld". Describe the fallback without spelling the legacy token.
Fixes the review findings on this PR that belong to files it owns, plus
several defects found while verifying those fixes. Findings in files owned
by upstream branches are routed there and left untouched here.
Correctness:
- tx_submitter: advance the account sequence only on results that actually
consume one (tes*, tec*, terQUEUED). tem*/tef*/tel* never reach the
ledger, so advancing left a permanent gap that every later submit from
that account inherited. Add a re-fetch hatch so a repeated non-consuming
failure cannot livelock on the same sequence, and gate the account check
on funded-ness rather than list length.
- validate_telemetry: filter spans by name before collecting attributes, so
a per-span attribute contract can no longer be satisfied by a sibling
span; require exact name equality for non-wildcard children and glob
matching for wildcards; bounds-check every returned series instead of
only the first.
- collect_system_metrics: select xrpld by argv[0] rather than a substring
match on the whole command line, which averaged in unrelated processes
and reported their RSS as xrpld's. Count genuine 0.0 CPU readings, use a
clamped nearest-rank p99 index, and record RPC latency only on success.
- benchmark: return each verdict through a named variable instead of a
command substitution, so the pass/fail counters survive and the exit gate
can fire. Scale before dividing in the percentage math, which truncated a
1.26% impact to 1.00% and cleared a 1% threshold.
- compare_to_baseline: fall back to the absolute bound when the baseline is
not positive, so a 0 -> 500 ms jump is no longer "within bounds".
- rpc_load_generator: bound each connection to one in-flight recv(), drain
in-flight requests before closing, use a nearest-rank percentile, and
report delivery shortfall so an under-delivered run cannot pass with a 0%
error rate.
Fail loudly instead of silently:
- run-full-validation: treat a consensus timeout and a missing validated
ledger as fatal infrastructure errors, and fold the orchestrator and
benchmark exit codes into the final status. A degraded cluster previously
ran a full validation pass and reported misleading downstream failures.
- collect_system_metrics: warn per empty measurement source, emit
metrics_complete, and exit non-zero instead of substituting zeros that
pass every threshold. Require GNU date with %N rather than falling back
to a per-sample python3 fork that costs more than the threshold it is
measured against.
- benchmark: distinguish "could not measure" from "exceeded thresholds",
install a cleanup trap so a failure cannot leak nodes and ports, and
report an unusable baseline as inconclusive.
- workload_orchestrator: bound subprocess communicate() and fail the exit
gate on per-phase errors.
Also pins the workload compose images to the versions the sibling stack
already uses, hash-pins the Python dependencies, restricts the validator
config template to loopback, corrects the dashboard and metric counts in
the reference docs, drops a span from the regression gate that cannot fire
under a WebSocket-only workload, and narrows the teardown pkill pattern so
it no longer matches processes that merely mention the work directory.
Verified with a full harness run against a local five-node cluster:
158 of 158 checks passed with no regressions detected.
The Telemetry Validation workflow failed with three "0 series" checks:
rpc_method_us, job_queued_us and job_running_us. All three are Histograms,
and the Prometheus exporter emits a histogram only as the
_bucket/_count/_sum triple -- the bare instrument name is never a series,
so validate_metrics() could never match it.
Evidence from the failing run (31804450127): its own metric-name dump
lists rpc_method_us_bucket/_count/_sum and no bare rpc_method_us, while
the sibling counters recorded in the same function bodies passed with 100
and 67 series. capture_timings.py, which queries job_queued_us_bucket and
job_running_us_bucket, returned real values for the acceptLedger job type
in that same run. Every one of the 10 histograms present exposes the full
triple, so all three suffixes are safe to assert.
Name them the way the exporter does, matching what the spanmetrics group
above already does for span_duration_milliseconds and what
regression-metrics.json and the job-queue dashboard already query. The
metrics stay in their asserted groups because they are genuinely
unconditional, so `not_asserted` would be wrong.
`.github/scripts/rename/docs.sh` is a migration script, not a check: the
`check-rename` workflow runs it and then fails if the tree changed. Its
rule `s@([^/+-])rippled@\1xrpld@g` rewrites `rippled` unless the preceding
character is `/`, `+` or `-`, so the backtick-quoted historical filenames
added by 3153f3ef56 (`` `rippled-*.json` ``, `` `rippled-alerts.yml` ``)
were rewritten and the check failed on both plan docs.
Refer to those historical names by their path-qualified form
(`dashboards/rippled-*`, `prometheus/rippled-alerts.yml`) instead. The
leading `/` exempts them from the rule, which is the form the rest of the
repo and the surrounding lines in these same files already use. No meaning
is lost -- the notes still record which commit renamed what.
tasks/fix-validation-checks.md was committed with the phase-10 harness and
shows up in the PR diff. tasks/ is a scratch area -- roughly forty analysis and
plan documents live there locally and none of the others are tracked, on this
branch or on develop. The file stays on disk; only the tracking is removed.
Every telemetry build instruction told the reader to run
`cmake --preset default`. No `CMakePresets.json` is tracked in the repo, and the
only preset that exists is `conan-release`, generated into the gitignored
`CMakeUserPresets.json` by `conan install`. So the documented first build step
fails before it starts, in the runbook's quick-start, its disable procedure, and
the telemetry stack's testing guide.
Replaced with the flow BUILD.md actually documents -- `conan install ..
--output-folder .` followed by an explicit `conan_toolchain.cmake` invocation --
with `-o telemetry=True` / `-Dtelemetry=ON` as the telemetry delta, and a note
that `--preset conan-release` works as a shorthand.
The harness manifests asserted things the code cannot produce and missed most
of what it does. Two assertions were failing every run, and the metric set
covered 16 of the ~41 emitted names.
expected_spans.json: rpc.process was required with rpc.ws_message as its
parent, but it is created only in ServerHandler::processRequest() on the HTTP
path, so a WebSocket-only workload never produces it -- it is now optional and
parented to rpc.http_request, and the rpc.process -> rpc.command.* edge is
skipped with the real reason instead of a coroutine-context-loss diagnosis that
was never the cause. Adds the missing rpc.ws_upgrade span, corrects four
parents (consensus.mode_change, pathfind.request, and update_positions/check,
which are children of consensus.establish rather than consensus.round), and
demotes conditionally-set attributes out of required_attributes so a healthy
run stops failing. Counts recomputed from the file: 41 span types, 62 unique
required attributes.
expected_metrics.json: 16 -> 52 asserted entries across the job-queue, RPC
method, reduce-relay, overflow and validation families, plus the fifteenth
dashboard uid. Metrics the harness workload cannot exercise -- erroring RPC,
ledger-mismatch, TxQ overflow, and the lazily-created getobject_* instruments
-- are listed in a not_asserted group the validator skips, rather than as
assertions that would fail on a healthy node.
The workflow's push trigger listed two globs matching nothing
(include/xrpl/basics/Telemetry*.h, src/xrpld/app/misc/Telemetry*), so no C++
telemetry change ever triggered validation. Replaced with the paths the code
actually lives in, including src/libxrpl/beast/insight/** for the insight
export path the harness depends on. The four inert workflow_dispatch inputs are
now labelled UNUSED rather than looking like working knobs.
Docs: the workload README described a StatsD dirty-flag mechanism under a
member name that does not exist, on a code path the harness never uses -- it
sets [insight] server=otel, so gauges export through an observable-gauge
callback every cycle. Adds the missing txq-burst phase, reconciles three
different dashboard counts, and drops "posts summary to PR", which the workflow
has no permission to do. The runbook's phase-10 section loses the last
sampling_ratio reference (not a config key), gains a Regression Gate and CI
subsection covering the gate that can fail CI, and its compose-logs command now
names the workload compose file. cmake --preset default is left for a separate
change: no CMakePresets.json is tracked, so it is wrong everywhere it appears.
Also drops the dead exporter=otlp_http key the harness wrote into every node
config, and stops capture_timings.py defaulting --profile to a profile that
does not exist.
Three doc conflicts, all inside Phase 10's own sections. Resolved by keeping
phase-9's corrections and re-applying phase-10's ownership of those sections on
top, rather than taking either side wholesale:
06 §6.8.3 — Status now says Phase 10 is implemented on THIS branch (phase-9's
copy said "not merged into this branch", true there, false here). The stale
"71 checks" enumeration and "Current Status" list are replaced by phase-9's
dynamic-count description plus a Known Gaps list that states the real reason
the rpc.process hierarchy is unassertable: rpc.process is HTTP-only, so a
WebSocket-only workload never produces it. Phase-9's CI Deliverable subsection
is carried through; exit criteria keep phase-10's tick state with phase-9's
corrected wording.
06 §6.8.3 Architecture — the automatic merge had resurrected phase-10's stale
2-node cluster prose and diagram over phase-9's 5-node correction. Reconciled
to phase-10's topology (native xrpld processes vs the containerised backend,
which is accurate) with phase-9's node count: 5 validators, and the collector
labelled OTLP + filelog rather than StatsD, which the config has never had.
Dropped the "all 26 metrics required" label in favour of the manifest.
06 §6.8.3 Key Implementation Details — two claims corrected against the code.
The StatsD m_dirty gauge fix describes a member that exists nowhere in the
repo, and the harness sets [insight] server=otel anyway, so gauges export
through an observable-gauge callback. The tx.receive attribute keys are bare
suppressed and tx_status, not dotted, and tx_status is set only on the
reject/known-bad/dropped paths, so it is absent on a successful receive.
09 §5c — kept phase-10's four-column table shape, with phase-9's corrected
counts: 40 of 41 emitted spans, 67 required attributes, 14 of 15 dashboards.
Phase10_taskList — dynamic inventory totals, the real RPC span trees, and exit
criteria ticked where the code on this branch closes them. Per-RPC timings are
recorded as not gated: regression-metrics.json defines only spans and job_queue.
Verified every hunk of the phase-10 diff falls inside a Phase-10-owned section
(06 §6.8.3, 09 §5c, Phase10_taskList, and the runbook's appended Phase 10
sections); no phase-9-owned text is modified from this branch.
The reference docs had drifted from the code in ways that break the reader
rather than merely misinform: PromQL examples that return no data, a rollback
flag that is a no-op, a sampling knob that does not exist, and two span parents
that moved. Code is treated as the truth throughout; where the code is the
defective side, the doc now records it as a known issue instead of describing
the bug as intent.
Renames the docs missed: histogram names gain the exporter's unit suffix
(ios_latency_milliseconds_bucket and four siblings), ledger_history_mismatch
gains _total, the StatsD-era quantile label gives way to le buckets,
rpc.request becomes rpc.http_request, traces_spanmetrics_calls_total becomes
span_calls_total, and the nine dotted xrpl.* span attributes are recorded as
renamed rather than left as live keys.
Re-parenting: consensus.update_positions and consensus.check are children of
consensus.establish, not of consensus.round.
Units and labels: state_accounting_*_duration is microseconds, not seconds;
cache_metrics label values are case-sensitive; object_count carries demangled
C++ type names. Nodestore read and write latency stays microseconds -- the
nanosecond accumulator change did not move the exported unit.
Adds what shipped but was undocumented: the ledger.acquire span, seven
consensus.round events, twelve span attributes, node_writes_duration_us, the
7-day validation-agreement window, the TxQ admission and reduce-relay metric
families, metrics_endpoint, and the phase-10 validation workflow.
Corrects claims that never held: 10% head sampling (it is fixed at 100%),
configurable redaction (it is unconditional), -DXRPL_ENABLE_TELEMETRY=OFF
(the flag is -Dtelemetry=OFF, default ON), FindOpenTelemetry.cmake and the
xrpl_telemetry target (neither exists), Promtail and a StatsD exporter in the
pipeline (neither exists), and Loki stream selection on job= (only
service_name is a stream label).
Phase 9 is marked complete, its provisioned alerting is attributed to the
branch that shipped it, and Phase 11 stays at zero except the one prerequisite
its code closes. Counts are reconciled repo-wide: 41 emitted span families,
15 dashboards on disk with 14 asserted, 13 alert rules in 5 groups.
Hardens the gate that let this drift through: Rule E of the naming check now
covers the reference docs, its allow-dotted marker is key-scoped and warns on
stale or empty use, a missing checked file is reported instead of silently
skipped, the test suite runs in CI, and doc paths trigger the check.
C++ and CMake changes are comment-only: three MetricsRegistry instrument names,
eight OTelCollector claims of a metric-name prefix that formatName never adds,
and the telemetry option's inverted default.
The Peer Disconnect Rate By Reason panel anchored its LogQL capture on
"\] ", which only matches a reason logged immediately after the [NNN]
peer-id prefix. PeerImp does not log that way: PeerImp::fail emits
"[NNN] <name> failed: <reason>" and the clean teardown emits
"close: Closed". Only ConnectAttempt::fail, which logs the bare reason,
ever matched. The panel's Timeout series was therefore connect-attempt
timeouts only, Ping Timeout was invisible, and PeerImp's own Closed was
uncounted.
Match all three prefixes and separate Ping Timeout from Connect Timeout.
Recorded as LogQL trap 11 in the runbook, alongside the other silent
failures this dashboard exposed.
Also document six overlay observability gaps found while auditing what
ping/pong and gossip traffic is actually tracked. All are pre-existing
and none is fixed here: the code fixes belong in develop-owned overlay
files (TrafficCount, OverlayImpl, PeerImp, PeerfinderManager), not on a
telemetry branch, and one of them needs a public signature change.
- 09 section 6: six known issues, each marked NOT IMPLEMENTED with
file:line evidence -- mtCLUSTER counted as unknown (overhead_cluster_*
always zero, 8 panels flatline), squelch_ignored byte counts always
zero, inbound/outbound byte-basis asymmetry plus a stale Total header
comment, ping/endpoints instrumentation absent, peer span coverage,
and PeerFinder exporting 2 of ~17 available readings.
- 02 section 2.3.2: add a Status column to the span catalog. Of 36
catalogued spans, 16 are live, 15 were never built, and 5 shipped
under different names (consensus.phase.establish became
consensus.establish, ledger.close became consensus.ledger_close,
rpc.request split into rpc.http_request/rpc.ws_message, txq.apply
became txq.apply_direct/txq.accept_tx). The catalog is a design
inventory; 09 section 1.1 remains authoritative for what emits.
- Phase9_taskList: tasks 9.14-9.17 tracking the deferred work, with
exit criteria checked only for what is actually done.
- Glossary: new Ping / pong keepalive term distinguishing ping timeout
from connect timeout. Correct the Cluster and Squelch entries, which
described behaviour the metrics cannot show.
The glossary header pointed at tasks/telemetry_terms.py as its
generator. That file is in no branch and nowhere on disk -- tasks/ is
gitignored one directory up -- so the header now states the file is
hand-maintained and gives the entry format.
Gates: check_otel_naming.py passes all 9 rules (Rule D over 555
dashboard queries, Rule E over the runbook); 19 doc anchors verified;
dashboard JSON valid with a one-line diff. No C++ changes.
The fetch and store duration counters converted each sample to
microseconds before adding it, so any backend call finishing in under a
microsecond contributed zero. A warm nudb read answers in a few hundred
nanoseconds, so on fast hardware every read floored and the totals stayed
at zero no matter how many reads happened -- the same loss of resolution
the microsecond report was introduced to avoid, one decade lower.
Both accumulators now hold nanoseconds, the clock's own resolution, and
convert once in getFetchDurationUs() and getStoreDurationUs(). The public
accessors, the node_reads_duration_us and node_writes_duration_us JSON
fields, and the metrics that read them all keep microseconds, so nothing
downstream changes unit. storeDurationStats() takes the raw duration
instead of a pre-converted integer so no caller can round early, and
updateFetchMetrics() scales its microsecond input to match.
FetchReport::elapsed stays microseconds: it carries one fetch, not a
total, and that is the unit it declares. The reported sum is therefore the
accumulated total minus a sub-microsecond remainder per fetch, so the two
tests that asserted exact equality between them now assert that bound.
Both had assertions that depended on how fast the host reads; the bound
holds on any hardware.
The debug-log notice on Log-Derived Insights was two grid rows taller than its
content needs, pushing the first data row further down than necessary. Rendered
the panel to confirm the markdown still fits with no clipping.
It is the first panel in the dashboard, so no other panel's stored y needed to
change - Grafana's vertical compaction closes the gap.
Carries the layout fix to the dashboards this branch owns. A repeating panel
claims the whole row once it expands per network, so the non-repeating panel
paired beside it was pushed down while keeping its stored x=12 - rendering on
the right with an empty gap on its left.
Repeating panels now get a row to themselves, keeping w=12 so their copies still
tile two across. Single-value panels are grouped to the top of each row section
so the charts that follow pair with each other rather than being separated by an
interleaved repeat.
Verified against origin/phase9: no panel lost, every targets block unchanged,
ids 1..N, no overlaps, no rows with a left-hand gap.
A repeating panel expands into one copy per network at view time and, with
maxPerRow=2, claims the whole row. The non-repeating panel paired beside it was
pushed to the next line but kept its stored x=12, so it rendered on the right
against an empty gap.
Two changes to how the layout is planned:
- a repeating panel gets a row to itself. It keeps w=12, so its copies still
tile two across inside that row.
- single-value panels are grouped to the top of each row section, so the
charts that follow pair with each other instead of being split up by an
interleaved repeat. Without this the gaps just become wasted half-rows.
Verified per dashboard: no panel lost, every targets block byte-identical, ids
1..N, no overlaps, and no row left with a gap on its left.
The GetObject Request Size Distribution heatmap, added on this branch, was
missing calculate, color and cellGap. Grafana's heatmap plugin needs them, and
without them the whole dashboard opens with "An error occurred within the
plugin" instead of rendering.
Same values as the other four heatmaps fixed upstream, taken from the one that
renders correctly on Grafana Cloud. The panel keeps its own axis label and unit.
Every heatmap carried only `tooltip` and `yAxis`, missing `calculate`, `color`
and `cellGap`. Grafana's heatmap plugin treats those as required, and without
them the panel fails to initialise: the dashboard opens with "An error occurred
within the plugin" rather than a chart.
The option values are taken from the one heatmap in this stack that does render
on Grafana Cloud (ledger-sync-health): calculate=false since the queries already
return histogram buckets, the Turbo 64-step scheme, and cellGap=1. Each panel
keeps its own yAxis label, unit and tooltip settings.
This is a long-standing defect rather than fallout from the recent layout work -
the same options are absent in origin/phase9 and in the cloud copies that were
already live.
Carries the sizing fix through to the dashboards this branch owns, after the
live Cloud copies showed charts clipping their legends at h=8 and stats too
cramped at h=4.
Every visualisation is h=10; tables and logs h=12; the log-derived-insights
instruction banner h=12 for its prose. A single height per visualisation also
means any two panels pair cleanly side by side.
Repeat stays on single-value panels only. The five state-timeline and heatmap
panels that had picked it up in an earlier pass are cleared, since repeating a
timeline broke ledger-data-sync on Cloud.
Verified against origin/phase9: no panel lost, every targets block unchanged,
ids 1..N, no grid overlaps.
Two problems showed up once these dashboards were live on Grafana Cloud.
Panels were too short. Charts at h=8 clipped their legends mid-row, and stats
at h=4 were cramped. Every visualisation is now h=10, with tables and logs at
h=12; the log-derived-insights instruction banner keeps h=12 for its prose.
Uniform height also means any two panels can pair side by side.
Repeat on a state-timeline broke the dashboard outright: ledger-data-sync
failed to open on Cloud at v45 and had to be restored to v44. Repeat is now
limited to single-value panels (stat, gauge, bargauge, table). Charts show
their networks as separate series instead, which is what a chart is for.
Repeat was also sticky: normalization only ever added the keys, so a panel that
picked them up in an earlier pass kept them even after its type stopped being
eligible. The keys and the title suffix are now removed from ineligible panels,
which is what actually cleared the two timeline panels here.
Verified per dashboard: no panel lost, every targets block byte-identical, ids
1..N, no grid overlaps, and repeat present only on stat/gauge/bargauge.
Applies the guideline-8 ordering to the two dashboards introduced on this
branch, and clears the last two labelling gaps found while auditing the set:
- peer-quality, validator-health: stats and bar gauges now lead their row
section instead of trailing the charts, matching the other dashboards.
- node-health "Peer Disconnects (Resources)": had an empty axisLabel, the
only chart in the set without one. Now reads "Disconnects".
- rpc-performance row "Aggregate RPC (all commands)" -> "(All Commands)",
the one title that was not in Title Case.
Reordering stays inside each row section so no panel changes category, and the
panel objects are cut and re-spliced as raw text, so their contents are
byte-identical. Verified against origin/phase9: no panel lost, every targets
block unchanged, ids 1..N, and every row section now leads with its
single-value panels.
Guideline 8 asks for gauges and stats at the top. Seven dashboards had them
scattered below charts, so the reader met a wall of time series before the
at-a-glance numbers that give those series context.
Stats, gauges and bar gauges now come first within each row section. The move
is deliberately scoped to inside a section: shifting a panel across a row
boundary would change which category it belongs to. Panels keep their relative
order otherwise, so the reading sequence within each group is unchanged.
Whole panel objects are cut and re-spliced as raw text, so their contents stay
byte-identical and only gridPos and id are recomputed. Verified per dashboard:
panel count unchanged, no panel lost, every targets block byte-identical, ids
still 1..N, and no row section left with a stat below a chart.
Applies the same normalization the phase-7-owned dashboards received, to the
five introduced on this branch: fee-market, job-queue, log-derived-insights,
peer-quality and validator-health.
- id written as 1..N so panelId deep links address a specific panel
rather than whichever panel happens to sit in that position
- gridPos at most two panels across; charts h=8, stats/gauges h=4,
tables/logs h=12 full width. Panels pair only with an
equal-height neighbour, so no row keeps a ragged empty cell.
validator-health and peer-quality were stacked single-file at
full width, which is why they were the longest to scroll.
- line charts lineWidth=1, fillOpacity=0, pointSize=5, gradientMode=none
- repeat xrpl_network_type (horizontal, maxPerRow=2) with a
[$xrpl_network_type] title suffix on stat/gauge/bargauge/
table/state-timeline panels, where two networks overlaid in
one panel reads as noise. Line charts keep their networks as
separate series.
- decimals 0 on panels counting discrete things (peers, quorum,
connection mix) - a fractional peer count is meaningless.
- rows category rows added to fee-market, job-queue and peer-quality
Panels with a right-hand legend stay full width; a side legend needs the room.
Raw-text edits, so untouched lines keep their formatting and escaping byte for
byte. Verified per dashboard against origin/phase9: panel count unchanged, no
panel lost, every targets block byte-identical, descriptions unchanged, ids
exactly 1..N, and no two panels overlapping on the grid.
