Two files conflicted and both were composed rather than taken from one side.
integration-test.sh: kept this branch's spanmetrics names, since the collector
sets namespace: "span" here and traces_span_metrics_* matches nothing, and took
phase-6's --max-time on every probe. The Tempo time bound needed restoring by
hand: RUN_START, the check_span guard and the start/end parameters are on
phase-6 and absent here, so a plain resolution kept phase-6's comment about
bounding the search while shipping no bound. All four pieces are back.
TESTING.md: kept server=otel with the metrics endpoint. Phase-6's template sets
server=statsd and documents prefix, which this branch's OTel path ignores.
This branch switches integration-test.sh to [insight] server=otel and adds an
assertion that no StatsD listener is needed, but TESTING.md still described the
metrics it verifies as StatsD-derived and its manual node-config template had no
[insight] stanza at all.
CollectorManagerImp falls through to NullCollector when server is neither statsd
nor otel, so a reader building configs from that template got zero
beast::insight metrics. The template also omitted service_instance_id and
metrics_endpoint, which the script writes; without the former every node is
indistinguishable in the $node dashboard filter.
curl applies no overall timeout of its own, so a server that accepts the
connection and then stops answering parks a poll loop for the rest of the run
and the loop's attempt count stops bounding anything. Add a CURL_MAX_TIME
ceiling and apply it to all 18 executable probes in integration-test.sh.
TESTING.md's manual node-config template also disagreed with what the script
writes, so a reader following it could not reproduce the automated path:
- no [insight] stanza, so no beast::insight metric leaves the node at all and
Step 10b's ten rippled_* assertions cannot pass
- [ips_fixed] listed all six peer ports including the node's own
The log level is deliberately untouched: the template and the script agree on
warning here.
check_statsd_metric queried rippled_rpc_requests, which no pipeline
produces: the collector's statsd receiver runs with is_monotonic_counter,
so the Prometheus exporter appends _total. A wrong name returns zero
series rather than an error, so the assertion could not be told apart
from a broken pipeline. All eight assertions were re-derived from how
each metric is created in code; this was the only counter.
Tempo searches carried no start/end, and tempo-data is a named volume
that `docker compose down` preserves under a one-hour block retention, so
the 17 span assertions could pass on an earlier local run's traces. Bound
every search to this run, and tear the stack down with -v before starting
so no earlier data is present to match. The service-name check now
matches a whole line, because the tag-values endpoint ignores start/end.
Add a gtest for the StatsD gauge that publishes its initial zero and for
the counter that must publish nothing. Assert two metrics the harness
never checked: a traffic-category gauge no message reaches, and
io_context latency.
Addresses review findings on the native-metrics work.
StatsDCollector::onTimer drained the send buffer inside the polling_ gate. That
gate holds back hook handlers until the application's services are built, but
sendBuffers() is socket I/O. StatsDEventImpl derives only from EventImpl, so it
never enters metrics_ and posts straight to the buffer; its |ms timings piled up
before onCollectionReady and were dropped after onCollectionStopping. The drain
now runs every tick, and outside metricsLock_, so onCollectionStopping no longer
waits on a UDP flush.
TelemetryImpl's constructor left meterProvider_ set when initMetrics() threw.
initMetrics publishes globally as its last step, so a throw left getMeter()
callers holding a provider nothing else could reach. Reset it in the catch.
~ApplicationImp caught only std::exception around telemetry shutdown while the
callees reach third-party SDK code, so a foreign exception would have terminated
the process. Added a logging catch-all.
ValidationTracker's hard trim evicted by unordered_map bucket order. It now
evicts oldest-first, so the entry dropped under pressure is the one least likely
to still reconcile.
The GetMeter test restored the global meter provider only on the success path,
and ASSERT_TRUE early-returns past it. Uses xrpl::ScopeExit instead.
The hook debounce window is a named constant rather than a bare 500 in a
comparison, and the metric export cadence becomes operator-configurable through
metric_export_interval_ms and metric_export_timeout_ms. Both are range-checked:
the SDK warns and silently substitutes its own 60s/30s defaults when the timeout
is not below the interval, so an unchecked value would slow export rather than
speed it up. Parsing uses a signed representation because lexical_cast<uint32_t>
accepts a leading minus and wraps it.
Naming corrections: CollectorManager documented exported_instance, which no OTel
dashboard uses; node-health queried job_count where the exported name is
jobq_job_count; network-traffic and overlay-traffic-detail referenced an
undeclared DS_PROMETHEUS variable; the counter table omitted the _total suffix
the Prometheus exporter appends; the plan docs and task list carried an xrpld_
prefix formatName never applies; and OTelCollector::New()'s contract promised its
instanceId, serviceName and networkType arguments were read, contradicting the
definition that marks them unused.
One [telemetry] key served both OTLP signals, and the metrics URL was
derived from it by suffix-swap: strip a trailing slash, strip a known
signal path if present, append the wanted one. Anything not ending
/v1/traces therefore posted metrics to the traces path, and the OTLP
version was pinned in code where an operator could not reach it.
Adds metrics_endpoint alongside traces_endpoint. Both are full URLs used
verbatim, so traces and metrics can go to different collectors, or to one
whose OTLP paths are not the defaults. signalEndpoint(), kTracesPath and
kMetricsPath are gone; nothing derives an endpoint from another.
The startup log names both URLs, since with two independent endpoints
there was otherwise no way to see where metrics were going.
Also drops exporter=otlp_http from the shipped config and the test
fixture. No branch in the chain reads an `exporter` key: it was a real
Setup member in the first phase-1b implementation, removed when only
OTLP/HTTP was wired up, and already deleted from TESTING.md once on the
same grounds.
Three conflicts, all composed rather than resolved by taking a side:
- TelemetryConfig.cpp: phase-6 kept networkTypeFromId file-local with
[[nodiscard]]; phase-7 had relocated it to public scope for
Application.cpp. Kept phase-7's relocation, so one definition remains.
The [[nodiscard]] survives on the declaration in Telemetry.h.
- Telemetry.cpp x2: phase-7 added getMeter overrides, phase-6 added
[[nodiscard]] to the startSpan below them. Kept both, and put
[[nodiscard]] on getMeter too.
- TESTING.md: phase-7 had the right metric name (span_calls_total, which
the spanmetrics namespace produces) but the wrong label. Its
xrpl.rpc.command appears nowhere else in the branch; the attribute is
bare `command`, which is what the dashboards query. Took phase-7's
metric with the correct label.
Both signalEndpoint call sites follow the renamed member. signalEndpoint
itself is left in place: removing it and adding metrics_endpoint is a
design change, not part of propagating a rename.
The rename arrived from phase-1b by merge. Four files still wrote the old
key, which the parser no longer reads, so each would have silently
fallen back to the default collector URL.
integration-test.sh is the load-bearing one: it generates the node config
the test harness starts, so the stale key would have pointed the node at
localhost regardless of the compose network. xrpld-telemetry.cfg is the
standalone node config; the other two document the key.
Note this cfg has a second, divergent variant on the devnet branches that
needs the same fix there.
[telemetry] endpoint carried one OTLP signal while its name implied it
covered every signal. That asymmetry is what let the metrics URL be
guessed later by rewriting this one's path suffix, so anything not
ending /v1/traces silently posted metrics to the traces path.
Renames the key to traces_endpoint and Setup::exporterEndpoint to
tracesEndpoint. The default value is unchanged and the URL is still used
verbatim, with no path derived from it. The startup log line and the
compose-file example name the new key, the latter being where an
operator copies it from.
No metrics_endpoint is added here: this branch has no metrics pipeline,
so the key would parse into a member nothing reads.
Harness and docs:
- integration-test.sh queried traces_span_metrics_* for spanmetrics, but this
branch sets the connector namespace to "span", so those two checks matched
nothing and failed. The dashboards and runbook had moved; the script had not.
- The same script queried eight native metric names with a product prefix and
capitals that formatName() cannot produce: it lowercases, maps '.' and ' ' to
'_', and prepends nothing. Corrected against the runbook tables.
- TESTING.md carried the same stale spanmetrics names and a jq example reading
a Prometheus label that does not exist.
- The runbook now records where each part of a derived metric name comes from,
since only the namespace is ours to choose.
Collector:
- OTelCounterImpl::increment silently dropped a negative amount. An OTel
counter takes unsigned deltas, so assert and let a release build under-count
rather than wrap.
- OTelGaugeImpl::increment computed current + amount in int64, which is
undefined on overflow, and the clamp ran afterwards so it could not help.
Check the headroom first. set() now clamps rather than casting a uint64 above
INT64_MAX to a negative, which is what made underflow reachable.
- The meter scope was two bare literals. They are constants now, and
Telemetry.cpp static_asserts them equal to kMeterName and kMeterVersion:
beast cannot include the telemetry header, so a build failure is the only way
to catch the copies drifting.
- formatName uses views::transform and ranges::to, as Backend.cpp already does.
- Unused constructor parameters take [[maybe_unused]] instead of (void) casts.
- The destructor logged "shutting down" and "stopped" with nothing between.
initMetrics was 79 lines doing four jobs. The exporter and the histogram views
are separate functions now, addUnitView is a member rather than a lambda
capturing this, and the export interval and timeout are named. It also derived
the metrics URL from the traces URL by suffix swap, which sent metrics to the
traces path whenever the configured URL had any other shape; both URLs now come
from one rule that handles a bare host, a trailing slash and either signal path.
Review feedback on the testing guide:
- rm -rf targeted data/, but this config writes under docker/telemetry/data/,
so teardown did nothing and a second run reused the old NuDB and SQLite
state. Corrected at both sites, including the Test 2 keygen node, which
launches with the same config.
- The standalone span table said consensus.* does not fire. It does:
ledger_accept drives a simulated round, so consensus.round, .phase.open,
.ledger_close, .accept and .accept.apply all appear. Only .establish,
.update_positions, .check, .proposal.* , .validation.receive and
.mode_change cannot. The test intro claimed the same thing and now agrees
with the table.
- Three blocks duplicated content the file already had. Test 1 now points at
the shared Verification Queries section as Test 2 already did, and the
Test 2 submit block checks engine_result like Test 1 does.
- The numbered step list was a copy of the script's own Step N headers and had
drifted by four entries, so it now points at those headers instead.
Also corrects the runbook's ledger and peer span tables against the code:
ledger.build was credited with tx_count and tx_failed, which tx.apply sets,
and was missing its three close-time attributes; peer.validation.receive was
missing ledger_hash and full_validation. The five source line numbers in those
two tables were stale, so they now name the file only, as the other nineteen
rows do.
Review feedback, plus a sweep of the branch for the same defects elsewhere.
Scope: the ledger.validate guard was a plain local, so it stayed alive until
checkAccept returned and the flag-ledger upgrade check ran inside the measured
span. That check reads every trusted validation of the parent, so one span in
256 became a duration outlier for work unrelated to promoting a ledger. The
span is now scoped to the promotion. tryAdvance stays inside it because it only
sets a flag and posts a job.
Attributes: tx.apply now carries ledger_seq, which the runbook already
documented. The parent ledger.build span has it, but a child cannot be selected
by its parent's attributes, so the span could not be found by ledger.
Guard names: each span guard is now named after the span it holds, so
proposalReceiveSpan, validationReceiveSpan, storeSpan and validateSpan. The
name "span" previously meant the trace root in one inbound-message handler and
the job-queue handle in its sibling, which taught a reader the opposite of the
truth in the next function.
Comments: the StatsD gauge rationale now sits with the initialiser it explains
rather than in the constructor. The peer span header described its trust flags
as shared when they are in fact re-declared to match the consensus keys; the
duplication is intentional and the wording was not.
Docs: the ledger and peer span tables disagreed with the code, crediting
ledger.build with attributes that are set on tx.apply and omitting several that
it does set, and all five source-file line numbers in them were stale. The
testing guide listed attribute keys that exist nowhere in the code, so its
catalog now points at the runbook instead of keeping a second copy that drifts.
OTelCollector routes every instrument name through a static formatName()
that only lowercases the name and maps '.' and space to '_'. The sole read
of prefix_ is the startup log line at OTelCollector.cpp:802, and all four
instrument factories go through formatName(), so no prefix can ever reach
an exported name. StatsDCollector does prepend it, so the StatsD example
keeps the key and now states why.
Covers the three server=otel blocks in the 09 reference and the config
integration-test.sh generates. This branch introduces OTelCollector, so it
is where the inert examples first appear; phase-6's examples are all
server=statsd and stay as they are.
formatName() never reads prefix, so setting it here does nothing and the
exported names are bare and lowercase. Leaving it invites queries written
against xrpld_jobq_job_count, which match no series.
The StatsD examples keep it, because that path does apply it to the name.
The panel described itself as showing which categories consume the most
receive bandwidth, and told operators to watch for a category dominating
during ongoing sync, but it ranked raw cumulative counters. That ranks by
lifetime total, so a sync burst days earlier still dominates the bars and
the stated guidance cannot work.
Wrapping the existing selector in rate() does not work: rate() drops
__name__, so the many *_bytes_in series collapse to one labelset and the
query fails with "vector cannot contain metrics with the same labelset".
Grouping with sum by (__name__) does not help either, because Prometheus
strips __name__ from aggregation output regardless of the grouping. Both
were tried against a local Prometheus with synthetic multi-category
counters and both error.
Instead recover the category into a normal `series` label before rating,
and use a subquery so the renamed series can be rated. Verified against
the same synthetic data: rates match a plain single-metric rate control
to within subquery resolution, ranking order is correct, and both the
category and node labels survive.
Also set unit to Bps, matching the ten sibling rate panels in this
dashboard (this was the only byte panel still on decbytes and the only
one plotting raw), and drop "Heatmap" from the title since the panel is
a bargauge.
Note the subquery evaluates the selector at several steps, so it costs
more than the previous instant query across the traffic categories.
This is the change that actually lifts the 5 s ceiling. Until now the
millisecond ladder and the Unit type existed but nothing consumed them.
Telemetry.cpp registered ONE histogram view: instrument name pattern "*",
unit exactly "ms", boundaries {1, 5, ..., 1000, 5000}. Verified against the
installed SDK, "*" matches every name and "ms" matches exactly, so that view
governed every beast::insight Event -- all 54 of them, whatever they measure.
Measured on devnet: 24.9% of rpc_size samples and 100% of jobq_updatepaths
samples fell above 5000. A quantile landing in the `+Inf` bucket reads back
as the second-highest edge, so those p95s reported a flat 5000 rather than a
measurement, and the 1 s to 5 s span was a single four-second-wide bucket
that any quantile inside it had to interpolate across.
Replaces it with one view per unit, keyed on the unit an instrument declares:
- `ms` gets kMillisecondBuckets: every representable edge of the collector's
spanmetrics ladder, plus 60 s and 120 s. The extensions are deliberate --
jobq_updatepaths was measured averaging 59,956 ms, which no span
approaches, so parity alone would still censor it.
- `By` gets kByteBuckets, placed from the measured response distribution
(mean 2131 B, half under 1 kB, tail mean bounded at 7538 B).
OTelEventImpl now derives its declared unit AND its description from unit()
instead of hardcoding "Duration in ms"/"ms", so rpc_size exports as
rpc_size_bytes on the byte ladder. rpc-pathfinding's "RPC Response Size"
panel follows the rename; its unit was already decbytes and is now truthful.
Also corrects Phase7_taskList.md, which still specified the 5000 ladder as
"matching SpanMetrics". That was true when written and became false when the
collector ladder was extended on its own -- implementing the plan as written
reproduced the bug, so the spec is where the defect had come to live. The
edges now have exactly one owner and the plan points at it.
beast::insight instruments are created during ApplicationImp's member-init
list, and opentelemetry-cpp 1.28 never rebinds an already-vended Meter, so an
instrument created before the MeterProvider is published records nothing for
the rest of the process. Observable instruments carry the opposite constraint:
registering one arms the SDK reader thread, and its callbacks run hook handlers
that read services which do not exist that early.
Publish the provider in Telemetry's constructor, ahead of every producer, and
defer only the observables. Collector gains onCollectionReady() and
onCollectionStopping(); OTelCollector arms and disarms its gauges in response.
StatsDCollector starts its polling thread in its own constructor and had the
same hazard, so it uses the pair to gate that thread.
The metrics resource carries service.instance.id and is immutable once built,
so the node public key is resolved in Main.cpp, where a config error can still
be reported, and passed to makeApplication(). getNodeIdentity() remains
authoritative; both paths now share readNodeIdentity(), so telemetry cannot
report a key the node has abandoned.
An explicit ~ApplicationImp stops observing and stops telemetry, covering the
setup() failure paths that never reach run(). Telemetry::stop() is once-only
and no longer clears another instance's global pointer. The histogram view's
meter selector now matches the meter actually in use, so its bucket boundaries
apply for the first time.
The transitions panel used increase(...[$__rate_interval]). $__rate_interval is
defined as max($__interval + scrape, 4 * scrape), i.e. deliberately one scrape
longer than the step so rate() windows overlap and lose no counter increase.
That overlap is harmless for rate(), but this panel reads the value as a count
of discrete events, and the overlap counts each event in more than one bucket.
Measured against a log-derived ground truth of 106 syncing transitions on
devnet-otel-usw2-01 over 2026-08-11T11:05Z..2026-08-12T23:04Z, the old query
reported 111.3 at a 300s step and 133.7 at a 60s step -- the error grew to +26%
as you zoomed in, because the overlap is a larger fraction of a smaller step.
Switch to $__interval so the buckets tile exactly, and wrap in round() because
increase() extrapolates to the window edges and so reports fractional counts for
an integer counter. The same measurement now gives 106 at 300s, 105 at 60s and
107 at 900s. Every state and both nodes land within a few counts of truth at any
zoom, and the legend Total is now a meaningful figure.
Pin Min step to 1m: the real scrape interval is 60s while the datasource
declares 15s, so without a floor $__interval can fall below one sample.
Draw as bars with 0 decimals -- the value is a discrete count per bucket, and a
line implies interpolation between counts that does not exist.
The Operating Mode Transitions panel queried state_accounting_*_transitions
directly. Those are monotonic counters, so the panel drew a slowly rising line
and a few transitions per hour were invisible against a total in the hundreds.
It also fell off a cliff whenever xrpld restarted and the counters reset to 0,
which reads as missing data rather than a restart.
Wrap each target in increase(...[$__rate_interval]) so each point is the number
of transitions in that bucket and the series survives a counter reset. This is
what the sibling panels on the same row (Operating Mode (Time Share), State
Duration Rate) already do.
Verified against devnet-otel-usw2-01/02 over 2026-08-11T11:01Z..2026-08-12T16:23Z:
the fixed expression reports 107 and 123 syncing transitions, matching the
counter deltas, and stays continuous across the 12:07 restart where the raw
counter dropped 630 -> 1.
Brief mode flaps remain invisible on Operating Mode (State Timeline) because a
~2 s dwell cannot be captured by a 60 s scrape; this panel is the place to read
them.
Conflict resolution kept this branch's evolution and re-applied phase-6's
fixes on top of it, rather than taking either side wholesale:
- consensus-health.json: kept the native `span_calls_total` metric name and
the `interval: 15s` and point styling from this branch; added phase-6's
`close_time_correct` PromQL filter and the NetClock axis labels. The
TraceQL boolean-regex filter stays removed and the `byRegexp` overrides
carry over. Panel count unchanged at 27.
- 09-data-collection-reference.md: kept this branch's headings, its more
detailed consensus attribute table (which already types
`consensus_round_id` as int64) and its section numbering, including the
deliberate removal of the SpanNames inventory. Carried over only the
correction that the state-accounting duration gauges are cumulative
microseconds, not seconds.
- telemetry-runbook.md: kept this branch's native metric names
(`span_calls_total`, `span_duration_milliseconds_bucket`); carried the
`rpc.request` -> `rpc.http_request` span-name fix and the `jobq_` segment
on the job-queue depth metric.
- integration-test.sh: kept this branch's `check_otel_metric` form and
carried the `jobq_job_count` correction.
The integration test asserted `rippled_job_count`, which never reports any
series, so that check always failed. `JobQueue` registers the gauge as
`makeGauge("job_count")`, but `Application.cpp` passes it
`collectorManager_->group("jobq")`, so the emitted StatsD name is
`jobq.job_count` and the exported Prometheus name is
`<prefix>_jobq_job_count`.
Corrected the same name in two runbook tables that also dropped the `jobq`
segment. `09-data-collection-reference.md` already had it right, which is why
the two documents disagreed.
Routed here rather than to the phase-10 PR where it was reported: the wrong
name is present in `integration-test.sh` on every branch from phase 6
onward, and this is the branch that introduces the file.
Left alone deliberately:
- `statsd-node-health.json` still queries the old name, but that dashboard is
deleted at phase 7 in favour of `node-health.json`
- `06-implementation-phases.md` names `job_count`, which is accurate as the
code-level makeGauge argument rather than the exported metric name
This branch removes the collector's StatsD receiver and un-publishes
8125/udp, but xrpld-telemetry.cfg still selected server=statsd, so the
sample config sent beast::insight metrics over UDP to a port nothing
listens on. Phase7_taskList.md:132 lists this switch as required work.
Select server=otel and replace address= with the OTLP metrics endpoint.
Document that endpoint and prefix are informational only, since
OTelCollector records on the global MeterProvider that [telemetry]
configures and formatName() applies no prefix, and note that beast
instruments are not exported yet because the collector is constructed
before the MeterProvider is registered.
All five TraceQL panels on this dashboard returned nothing, and did so
without any visible error: they filtered on
span.close_time_correct=~"$close_time_correct", but close_time_correct is
a boolean attribute (RCLConsensus.cpp:601 passes a raw bool), and Tempo
restricts the regex operator to string operands, so the spanset resolved
to false. With the variable defaulting to All the clause rendered as
=~".*", so the panels were empty out of the box and looked exactly like a
node with no consensus activity.
Note this is the opposite of PromQL, where an absent or empty label does
match ".*" — which is why the 17 Prometheus panels on this same board were
unaffected and the dashboard appeared healthy.
Dropped the clause from all six queries, matching phases 9 and 10 where it
is already gone. The $close_time_correct variable now filters the
Prometheus "Close Time Agreement" panel instead, which already grouped by
that label but never filtered on it, so the control stays useful rather
than becoming dead UI.
Two defects were masked behind the empty panels and are fixed too:
- "Close Time: Raw Proposals" and "Close Time: Effective / Quantized"
carried unit dateTimeFromNow over close_time_self/close_time, which are
NetClock seconds (Ripple epoch), while Grafana's dateTime formatters
expect a millisecond Unix epoch — every point would have rendered as
roughly 1970. They now plot as plain numbers with the axis labelled
"NetClock Seconds (Ripple Epoch)", and the descriptions give the
946684800 offset for converting to Unix time.
- "Close Time Vote Bins & Resolution" matched its unit and axis overrides
byName against "Vote Bins" and "Resolution", which are not field names;
TraceQL select() yields close_time_vote_bins and close_resolution_ms, so
neither override applied. Switched to byRegexp so the match holds
whichever scope prefix Grafana emits.
No panel was added or removed: the (type, title) multiset is unchanged at
22. resolution_direction keeps its regex filter, which is correct there —
it is set from a std::string whose values are exactly the variable's
increased/decreased/unchanged.
The integration test's span assertions never actually ran. check_span()
built a Tempo /api/search call with --data-urlencode but no -G, so curl
POSTed the params as a body; Tempo answers 200 and ignores the query, so
every span name looked present. Verified against a live Tempo 2.9.4: the
buggy form returns the store's total trace count for any name, including
"zzz.does.not.exist"; with -G a real name returns 1 and a bogus one 0.
Fixed alongside it: the RPC check asserted "rpc.request", which is never
emitted (ServerHandler.cpp builds "rpc.http_request"). These two had to
change together, since -G turns the bogus name from a silent pass into a
hard failure.
Also in the script: a consensus timeout logged two failures and counted
two, because a post-loop else re-reported what the timeout branch had
already reported; and three unguarded curl calls aborted the whole script
under set -euo pipefail, making the ACCOUNT_ZERO fallback dead code with
no cleanup. Guarded the curls and wired an EXIT trap to the existing
cleanup(). The trap deliberately fires only before the summary, so a
completed run still leaves the stack up as the header documents.
Docs corrections, all re-derived from code:
- span inventory heading 35 -> 38, attribute heading 83 -> 89 rows
(78 unique keys), and the section 6 header table now carries the
missing TxApplySpanNames.h row so its columns sum to the same figures
- two stale paths: ConsensusSpanNames.h is under include/xrpl/consensus/,
TxSpanNames.h under src/xrpld/telemetry/
- consensus_round_id is int64, not string (RCLConsensus.cpp sets
prevLgr.seq() + 1); the runbook's TraceQL examples now use a numeric
literal instead of an unparseable bare <round_id>
- state-accounting duration gauges are cumulative MICROSECONDS, not
seconds (NetworkOPs.cpp declares std::chrono::microseconds and
publishes dur.count() raw)
- sampling_ratio is not a config key; head sampling is fixed at 1.0 and
the shipped collector has no tail sampling, so the caveat was rewritten
- the plan blurb referenced Jaeger; this stack is Tempo
Run with no arguments the script iterated an empty list, found no
violations and printed "OK: 0 dashboard(s) passed" with exit 0 -- a clean
bill of health for reading no files, indistinguishable from a real pass.
A bare run now defaults to every dashboard beside the script, and a run
that still ends up with nothing to check exits 2 rather than reporting
success. Passing paths explicitly behaves as before.
Nine targets across four panels filtered on xrpl_work_item while leaving
it out of their sum by() grouping. PromQL keeps only the labels listed in
by(), so the label was dropped from the result and the xrpl_ident legend
these panels build from it lost that segment.
Perf-iac runs stamp xrpl_work_item and give each work item its own set of
nodes, so service_instance_id already separates the runs; the visible
effect is the legend rather than merged series. Adding the label makes the
work item identifiable, which is the comparison these panels exist for.
NuDB Read Latency and NuDB Read Found Ratio each divide two aggregations.
Both sides get the label so their label sets stay equal and vector
matching still works.
The row was titled after the rollout phase its metrics came from, a number
defined only in a planning folder outside the shipped tree. Grafana shows
this title to operators, so it named something no reader could look up.
Two panel descriptions pointed at a rollout phase and task number defined
only in a planning folder outside the shipped tree. The note still names
the file and the change it is waiting on, which is the part a reader can
act on.
The catalog tagged every span with a rollout phase number defined only in
a planning folder outside the shipped tree, so the column meant nothing
to a reader of the repository. The span name and source file identify
each entry.
The section labels carried rollout phase numbers that are defined only in
a planning folder outside the shipped tree, so they meant nothing to a
reader of the repository. The descriptive half of each label already
identifies the section.
The label carried a rollout phase number defined only in a planning
folder outside the shipped tree. The descriptive half already identifies
the section.
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.
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.
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.
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.
These dashboards were hand-authored over time and had drifted apart: panel
heights spanned ten different values, nine dashboards had no row grouping,
line-chart styling was inconsistent, and no panel carried an id, so Grafana
assigned them positionally at load and every panelId deep link was only as
stable as the panel order.
Per panel, in document order:
- id written as 1..N so panelId links address a specific panel
- gridPos quantized to at most two panels across: charts h=8,
stats/gauges h=4, tables/logs h=12 full width. Panels are
paired only with an equal-height neighbour, so no row is left
with a ragged half-empty cell.
- 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 the panel types where
overlaying two networks in one panel reads as noise
(stat/gauge/bargauge/table/state-timeline). Line charts keep
their networks as separate series, which is the point of a
line chart.
- decimals 0 where the value counts discrete things (threads, peers,
queue depths); a fractional thread count is meaningless.
- rows category rows added where a dashboard had none
Panels whose legend sits on the right stay full width: a side legend needs the
horizontal room, and squeezing it to half width clips the series names.
Edits were made as raw-text replacements, not a json.dump round-trip, so
formatting and escaping of untouched lines are byte-identical. Verified per
dashboard: panel count unchanged, every targets block byte-identical, all
descriptions unchanged, ids exactly 1..N, and no two panels overlapping on the
grid. The repo dashboard lint and the OTel naming check both pass.
The Transaction Overview panel "Queue Bypass Ratio (Direct Apply vs
Enqueue)" reported a confident 0.50 on every node while the true bypass
rate was zero. The two spans it divided are not disjoint alternatives:
txq.apply_direct is a child of txq.enqueue. TxQ.cpp creates the
apply_direct span as the first statement of tryDirectApply(), ahead of
the account, sequence and fee-level guards, and tryDirectApply() is
called from inside the live enqueue scope. The span therefore counts
attempts, so the denominator direct + enqueue counts each transaction
twice and pins the ratio to one half algebraically.
Measured on a four-node fleet: 6082443 direct against 6082877 enqueue
over the same population, panel output 0.5000170 on three nodes and
0.5000000 on the fourth. Grouping txq.enqueue by txq_status over seven
days returns only "rejected" -- no transaction has ever taken the
direct-apply path.
Remove the panel rather than repoint it. A correct expression using
txq_status as the disjoint discriminator would render permanently
empty on this fleet, which reads no better than a wrong number.
Widen the band partner "TxQ Enqueue Rate by Transaction Type" from 12
to 24 columns so the y=48 band still fills the grid. Every band in all
ten dashboards sums to 24 columns; leaving a half-width hole would be
the only exception. Panel order and every other panel's position,
width and height are unchanged.
The runbook already listed txq.apply_direct as available but not
paneled, so that row becomes accurate. Rows describing the span itself
are untouched -- the span and its metric are unchanged.
Panels 21 (NuDB Read Latency) and 23 (NuDB Read Found Ratio) on the
Ledger Data & Sync dashboard guarded their divisor with
clamp_min(<denominator>, 1). clamp_min raises the value, not just the
zero case, so any node reading fewer than 1 block per second was
divided by a fabricated 1 instead of by its real read count.
Replace the clamp with the filter (<denominator> > 0). A comparison
without the bool modifier drops the sample rather than rewriting it, so
these panels now show no data instead of a wrong number.
Measured over 7 days: five nodes fall below 1 read/s. On validator-0 the
clamp reported 2.726 us/read against a true 5.493, and on nonUNLmalloc-tc
it reported 0 us/read, which cannot occur. The error is largest exactly
when panel 21 is used as the bottleneck discriminator during a stall,
because that is when the read rate collapses toward zero.
Matches the existing idiom on the same nodestore_state metric family in
the NodeStore Write vs Read Latency panel.
