* fix(servers): compose OTLP metrics job from service.namespace/service.name
The OTel Prometheus compatibility spec defines job as
"<service.namespace>/<service.name>" when the namespace is present.
The OTLP metrics path only used the bare service.name, so the job tag
diverged from target_info produced by Prometheus-side exporters for the
same resource. Compose the namespace form, and keep not fabricating a
job when service.name is absent.
Behavior change: resources carrying service.namespace now get
"namespace/name" as their job tag value.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* feat(otlp): synthesize otel_resource_info at OTLP metrics ingestion
Ordinary OTLP metrics scatter filtered resource attributes as tags over
every logical metric table, so metrics-only services contribute nothing
to the semantic entity graph. Each request now also projects its
distinct resources into one info-metric-shaped mito table,
otel_resource_info: a fixed allowlist of identity-relevant attributes
under their raw OTel keys (independent of the label translation
strategy and the promote/ignore headers) plus derived job/instance
compatibility columns, value 1.0, and the newest data-point timestamp.
The descriptor is written after the main insert is committed; a failure
there (conflicting pre-existing table, auto-create disabled) degrades
to an OTLP partial_success warning with rejected_data_points = 0
instead of failing the request and triggering client retries of
already-accepted data. A request writing a metric named
otel_resource_info suppresses synthesis. Legacy mode is unchanged.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* feat(operator): otel info-metric conventions with host/container entities
Whitelist the ingestion-synthesized otel_resource_info descriptor via a
new otel_info_metrics conventions map, gated on source=opentelemetry
(the existing gate hardcoded source=prometheus). Its declarations use
explicit descriptive lists instead of descriptive_rest so identifying
attributes of other entities do not leak into service.instance.
Conventions tightened per the Astronomy Shop findings: host identity is
host.id with host.name descriptive only (host.name is not stable across
SDKs and resource detectors), a generic container entity (new entity
type) is declared only when container.id is present, and trace-v1
tables now synthesize host/container from their flattened resource
attributes too. New co-declared edges: service.instance runs_on
container, container runs_on host.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* test(otlp): cover the resource descriptor in integration tests
Covers the descriptor's raw-key columns and info-metric options through
the HTTP path, the namespace/name job composition end-to-end, column
names being independent of the translation strategy, the allowlist
excluding unlisted resource attributes, auto-create after a drop, the
metric-name collision suppressing synthesis, and the partial-success
warning (rejected_data_points = 0) when a pre-existing incompatible
table fails the descriptor write while metric data is accepted.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* chore: cargo fmt
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* fix(frontend): degrade descriptor permission denial to a warning
A table-level permission policy denying otel_resource_info would have
failed the whole OTLP metrics request because the descriptor's
permission check ran before the main insert. The descriptor is derived
enrichment: check its permission in the degrade path so a denial skips
the write and surfaces as the partial-success warning, like any other
descriptor write failure.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* fix(otlp): guard descriptor writes with semantic ownership markers
A pre-existing schema-compatible table named otel_resource_info would
silently receive descriptor rows while its missing semantic stamps kept
it out of the entity graph. The descriptor write now requires the
auto-created table's ownership markers (mito engine + signal_type +
source + metric.type=info + metadata_quality=declared) and otherwise
degrades to the partial-success warning; the entity-graph gate for the
otel whitelist likewise requires metric.type=info, so a user table
stamped with only signal/source no longer picks up implicit
declarations.
Also fold the descriptor write cost into the response and surface the
degrade warning through the otel-arrow BatchStatus status_message.
Integration tests pin the full marker set on auto-create and that an
existing owned descriptor keeps accepting writes without degrading —
a missing marker would otherwise silently stop every descriptor write
after the first request.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* perf(otlp): build descriptor rows without the per-resource BTreeMap
Projecting a resource allocated a BTreeMap and then collected it into the
row key, and every attribute was matched against the allowlist by linear
scan. Collect the tags into a Vec and sort once, and match the allowlist
instead of scanning it. Measured on the conversion path: descriptor work
drops 16-18%, from 10.6% to 8.9% of conversion CPU on the worst shape
(1000 resources with 4 data points each), where the cost tracks resource
count rather than data-point count.
Also trims the comments and tests added with the descriptor to what
carries information.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* test(otlp): pin the descriptor permission-denial degrade path
A policy denying the descriptor table must not fail the metrics request,
which the fix in 2401b3dd9c does but nothing covered. Verified as a
regression guard by mutation: moving the permission check back before
the main write makes this test fail.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* test(otlp): keep legacy mode covered after trimming the unit tests
Trimming the descriptor tests dropped the only assertion that legacy
mode skips the job/instance remap and the promote filter. Both alter
the columns of tables already in use, so fold the check into the legacy
conversion test rather than leaving it uncovered.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* fix(semantic-graph): stop encoding column names into composite entity ids
A composite entity id rendered the identifying columns as sorted
`col=value` pairs, so the same identity split into one entity per
signal: a trace table names its columns service_name and
resource_attributes.service.instance.id where a metric table names them
job and instance. One service instance became two nodes with two
parallel edge sets, breaking the walk from a trace to that instance's
metrics.
Render an id as its values in declared order instead, escaping the
separator so components stay distinguishable, which is what single-column
ids already did by keeping only the value. entity_id_attrs still carries
the structured form.
Values alone are not enough for a namespaced service: the metric side
folds service.namespace into job while traces keep the bare name. Add
qualified_by to the conventions so the trace declarations compose the
namespace the same way, per the OTel rule that job is
<service.namespace>/<service.name> or the bare name when the namespace
is empty. A table without the namespace column keeps the unqualified
identity rather than losing the declaration.
Conventions validation now rejects one entity type declared with a
different number of id columns by two sources, which would silently
produce ids that can never match.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* style(otlp): import the parent module by crate path
check-super-imports.py, part of the CI format gate, rejects a
file-level `use super::`.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* feat(otlp): gate the resource descriptor, and fix what review found
Synthesizing greptime_otel_resource_info creates and writes a table the
user never sent, so it is now off unless
otlp.experimental_enable_resource_info says otherwise. With it off the
request costs exactly what it did before the descriptor existed: nothing
is projected, no table is created, no write and no permission check
happen. Tests run with it on. StandaloneOptions carried no otlp field,
so the whole [otlp] section was silently dropped in standalone mode; map
it through, or the new option (and trace_ingest_chunk_size before it)
would do nothing there.
Renamed from otel_resource_info: the greptime_ prefix marks the table as
engine-managed and makes a collision with a user metric unlikely, which
is what the pre-existing-table ownership check and its per-request
catalog lookup were defending against. Both are gone.
A request may carry data for several graph windows, but the descriptor
folded every data-point time into one row at the newest of them, leaving
the earlier windows with metric rows and no entities. Key the rows by
window as well, and take the times from the data points the encoder
actually writes: it drops exponential histograms, and a resource
carrying nothing else was being described as an entity with no
measurements.
Projecting a resource cloned its attributes once per data point. Nest
the windows under the attributes instead, so they are moved once per
resource, and walk the data-point times through a visitor rather than
collecting a Vec per metric.
Also documents what the two maps key and hold, and lifts the projected
attribute names to constants beside KEY_SERVICE_NAME.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* fix(otlp): skip the descriptor's work entirely when it is disabled
The collision scan over the request's output tables ran even with the
feature off. Short-circuit on the option instead, and update the config
snapshot the new [otlp] section changed.
Also drops the doc comment orphaned by the deleted ownership check: it
had attached itself to the trait impl and described a check that no
longer exists.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* refactor(semantic-graph): drop the expect and name the service identity
The CASE is built through Case directly rather than the fallible
when().otherwise() builder, so the non-test path no longer carries an
expect (architecture-invariants $4).
service_identity returned two same-typed Options that both call sites
destructured positionally; a named struct makes a swap fail to compile.
Also records that id-column order is part of the identity, where the
option docs and the conventions authors will read it.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* chore(semantic-graph): drop comments that narrate the code
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* fix(semantic-graph): cast duration_nano before the trace-table union
Trace tables written before the signed-integer ingest change hold
duration_nano as UInt64 and later ones as Int64. The calls derivation
unions the per-table selects, and the two have no common integer type,
so a deployment holding both shapes could not build the plan. The
cross-table test now spans both.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* refactor(semantic-graph): drop the redundant duration_nano casts
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* fix(otlp): decide exponential histogram acceptance in one place
The resource descriptor mirrored only the experimental gate, so with both
experimental flags on a resource whose only metric is a delta exponential
histogram was described as an entity with no measurements. The encoder's
whole-metric rules move into exponential_histogram_gate, which both call,
and the descriptor takes its timestamps through exponential_histogram_value
so per-point rejections drop out too.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
---------
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: add entity relationships and graph query RFC
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: revise entity-graph RFC after review
- Zero-configuration declarations: the Prometheus-on-Kubernetes convention
pack (job/instance per the OTel-Prometheus compatibility spec, k8s SD
labels, *_info descriptors, target_info enrichment) alongside the OTLP
trace auto-stamp, plus Remote Write 2.0 inline metadata.
- Calls endpoints follow the service entity declaration; self-calls
compare full endpoint ids; no silent identity fallback.
- Strict time-window contract: the source window is never narrower than
the query's observed_at range; unsafe-to-extract predicates error
instead of silently defaulting.
- scope removed from the relationship schema (kept on entities as a
display property); entity row contract restated per projected
observation; endpoint encoding documented as the v1 storage-level key
with its known collision limitation.
- Sampling caveats corrected (ratios are representative only under
unbiased sampling); snapshot relation synthesizes endpoint-only
vertices; shared attributes provide join keys while co-declaration
provides relationship semantics.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: align entity-graph RFC contracts and tighten prose
- Metric/trace service unification is promised only when service.namespace
is empty and job is not relabeled (the compatibility spec renders job as
<namespace>/<name>); otherwise alignment needs pipeline normalization or
explicit declarations.
- Entity row contract stated once (per projected observation); the calls
defining SQL is marked as the single-column simplification of the
declaration-derived endpoint ids; Remote Write 2.0 metadata intake and
the Prometheus implicit declarations are listed as M1 work.
- Compress survey/example/reference prose.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: pin down the k8s convention pack rules and edge directions
- kube_pod_owner implicitly declares k8s.workload with id
(namespace, owner_kind, owner_name) and derives pod part_of workload;
target_info's non-job/instance labels are implicit service descriptive
columns — fixed rules, no new declaration syntax.
- One direction for pod placement: k8s.pod runs_on k8s.node (pod added to
runs_on sources; node->pod removed from contains).
- Drop the remaining 'canonical' wording for the v1 storage-level id.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: make contains/part_of a true inverse pair
part_of covers service.instance->service and k8s.pod->k8s.workload with
contains as its inverse; has_instance is dropped.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: rewrite entity-graph RFC as a design narrative
Restructure for human review: mainline first, cases illustrate the
design instead of specifying it.
- New Architecture and Benefits and Drawbacks sections; the calls
derivation stays expanded as the flagship example while schema
enumerations, window-rule listings, and executor edge-case handling
move out of the document.
- The cross-signal promise is stated honestly: neighbours and their
source tables are discovered first, their telemetry is the next
query — one engine, two statements; the worked example shows the
full declaration -> entity -> edge -> telemetry flow.
- Default materialisation added as the most direct alternative to
read-time derivation, with its costs.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: final wording pass on the entity-graph RFC
Mark the property-graph DDL as illustrative rather than settled M2
syntax, credit standards as foundations rather than claiming wholesale
alignment, and clean up punctuation-heavy prose.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: keep the RFC at design altitude
Demote the convention-pack rule details and snapshot property-merge
semantics to the implementing changes; correct the single-trace-table
assumption (traces can be routed to multiple tables); record
attribute-key participation in entity equality as an open question.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* docs: correct service graph terminology
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
---------
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* feat: table semantic layer information_schema view (Phase 3)
Add `information_schema.table_semantics`, a queryable view over the table
semantic layer. One row per table that carries at least one
`greptime.semantic.*` option: the signal-agnostic keys
(signal_type/source/pipeline/metadata_quality) are promoted to columns and
the remaining signal-specific keys are folded into a `semantic_options`
JSON string. Tables with no semantic key are excluded.
Stacked on Phase 2.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* chore: address PR review on table_semantics
- fold JSON serialization failure into None instead of unwrap/panic
- drop per-row Vec allocation in predicate eval; use a fixed array
- align RFC view name with the shipped `table_semantics`
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* chore: update results
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
---------
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* feat: table semantic layer per-table enrichment (Phase 2)
Phase 2 of the table semantic layer, plus a vocabulary trim so the layer only
records what a machine consumer cannot cheaply recover on its own.
Per-table metric enrichment (OTLP), via an internal per-table channel:
- A `SemanticIndex` accumulator records, per emitted table, the declared metric
keys: type / unit / temporality / metadata_quality=declared / original_name.
Conflicting single-valued keys collapse to `mixed`/`unknown`.
- Recording happens at the `encode_metrics` level where the base name, metric
type, and proto fields are all in scope, so histogram/summary fan-out gets the
correct per-subtable type (`_bucket`=histogram, `_sum`/`_count`=counter)
without threading state through every encoder.
- The index is serialized onto the `greptime.internal.semantic.per_table_index`
context extension; `apply_per_table_semantic_options` folds each table's keys
into its options at auto-create time.
- `trace.conventions` is refined from the request's resource/scope `schema_url`s
(concrete when uniform, else `mixed`/`unknown`).
Vocabulary trimmed to only meaningful keys. Kept: signal_type, source, pipeline,
trace.conventions, metric.{type,unit,temporality,metadata_quality,original_name}.
Dropped: metric.monotonic (a function of type), trace.has_events/has_links
(constant + derivable from columns), log.severity_scheme/body_format (constant /
derivable, and body_format cost an O(rows) scan), resource/scope lineage
(restates columns / collector-config concern), source_version (no cheap
non-constant value today). Prometheus carries type/unit in the metric name by
convention, so it gets identity only — no inferred enrichment.
Identity (signal_type + source) extended to the remaining ingest protocols so
the discovery view is complete: InfluxDB and OpenTSDB (metric), Loki and
Elasticsearch (log). These protocols carry no type/unit metadata, so identity is
all that applies.
Tests: unit coverage for the accumulator, per-metric-type fan-out, and trace
conventions; integration goldens updated for the OTLP metric/trace SHOW CREATE
output and the new Loki identity.
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* chore: validate the option value
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
---------
Signed-off-by: Dennis Zhuang <killme2008@gmail.com>
* feat: Runner executes procedure
* feat: Add rollback key type to ParsedKey
* feat: Write rollback key when procedure is unable to execute
* feat: Use loaded step to re-submit subprocedure
* feat: Track subprocedures in ProcedureMeta
* feat: Clean message cache after the root procedure is done
* feat: Runner returns execution result
* fix: Fix tests
* test: Test Runner
* test: Test procedures_in_tree
* chore: Refine test and comments
* feat: Remove support of lock inheritance
A deadlock happens if a subprocedure acquires the same lock key as
its parent.
The main concern is if the subprocedure directly inherits its parent's
lock, then how should we behave when multiple subprocedures acquire
this same lock? Each procedure may assume it has unique access to the
same object but it actually shares the resource with others.
Now subprocedures need to use different keys to lock objects, which is
reasonable. For example:
- A parent procedure wants to create a table so it locks the table with
a key like `catalog.schema.table`
- Subprocedures create regions for the table so they lock the regions
with keys `catalog.schema.table.region-0 ~ catalog.schema.table.region-n`
* style: Fix clippy
* feat: insert_procedure returns false on duplicate procedure
Also rename this method to try_insert_procedure
* chore: Address CR comments
* docs: Add procedure framework RFC
* docs: Add dump, rollback and locking to procedure framework
* docs: Change ProcedureBuilder to ProcedureLoader
* docs: Add sub-procedures section
* docs: Add a link to explain idempotent
* docs: Add link to the tracking issue
* docs: Fix ProcedureLoader type alias
* docs: Update procedure API
* docs: Address CR comments
* docs: Update path and make the docs more clear
dennis zhuang
shuiyisong
discord9
jeremyhi
Sicong Hu
yihong
Yuhan Wang
Yohan Wal
gcmutator
tison
Wei
Zhenchi
zyy17
Lei, HUANG
LFC
Zhizhen He
elijah