Files
clean-architecture-backend-…/docs/registries/metrics.yaml
T
DongHyeonka bbe944bd3c fix!: 재리뷰가 찾은 결함과 색인에서 빠졌던 발견 5건을 고친다
문제: 성숙도 리뷰 수정을 다시 읽은 재리뷰(Critical 0, 회귀 0)가
Important 27건·Minor 41건을 찾았다. 수정이 새로 만든 결함이 많았다.
@SystemInvoked가 권한 규칙의 탈출구가 됐고, 데모 계정 거부가 정확히
prod일 때만 걸렸고, outbox 순서 번호가 업그레이드 때 뒤로 갔고,
GraphQL union 아래 복잡도가 0이었고, 서블릿 오류 응답이 success:true로
나갔고, 호출자 deadline이 breaker를 열었다. 원래 리뷰 색인이 06 영역
발견 5건(06-10~14)을 빠뜨려 그동안 처리되지 않았다.

방법: 영역별 7묶음이 test-driven-development 절차로 실패 테스트를
먼저 쓰고 고쳤다. 결정은 재리뷰 README 4절에 적었다: 순서 번호는 V21이
기록한 전역 floor부터, 데모 계정은 local·test·demo 허용 목록,
@SystemInvoked bean은 기동 실패, complete 단계 실패는
IDEMPOTENCY_COMPLETION_UNKNOWN(결과 불명), 캐시 조건부 쓰기는 별도
SCRIPT lane, strict lane 정책은 build cache 입력, 끝나지 않는 재시도
행은 원인 코드·지표·PARKED 절차. 레지스트리가 선언만 하고 아무 코드도
내보내지 않던 redis.capability.* 지표 6개는 빼고 runbook을 실제
지표(redis.lane.*) 기준으로 고쳤다.

위험·한계: V21을 V22 없이 직접 고쳤다(어느 배포에도 아직 없음). 이
브랜치의 V21을 적용한 개발 DB는 다시 만들어야 한다. 서블릿 오류
디스패치를 보안 체인이 허용한다. 발행·저장소 포트 시그니처가 바뀌어
파생 서비스 코드가 깨진다. prod에서 fixture-로 시작하는 설정값은
기동을 막는다. httpclient failure-injection lane의 main 단계 중복
실행은 이전 결정대로 남겼다. 실제 CI runner 실행은 확인하지 못했다.

검증: ci, mainCheck, :integrationCheck, -p optional-platforms ci
integrationCheck performanceCheck, :performanceCheck, :releaseCheck
(-PgitRevision), verify-compose-profile-contracts(18 lane) 통과.
첫 실행의 실패 2건(FindSecBugs 1, sample 동시 생성 테스트의 v4 event
id)은 고친 뒤 ci·integrationCheck·mainCheck를 다시 돌려 통과했다.

Refs: docs/reviews/2026-11-03-rereview/README.md 3·4·6·7절
BREAKING CHANGE: outbox 발행 포트와 저장소 포트 시그니처, 순서 번호
첫 값(V21 floor 다음), 샘플 데모 계정 프로필 허용 목록, 새 오류 코드
IDEMPOTENCY_COMPLETION_UNKNOWN, prod의 fixture- 값 거부, 기동 rule은
static·의존 없음이어야 한다.
2026-11-03 23:29:35 +09:00

760 lines
28 KiB
YAML

# Registry: Metrics
# Schema owner: feature-contract-registry-governance
# Owner branch: feature-metrics-alerting-contract
# Last updated: 2026-05-22
#
# Notes
# - Naming: Micrometer dot.case + unit suffix (.seconds | .bytes | .total).
# - Tag cardinality bounds are SSOT of feature-metrics-alerting-contract "Cardinality Bounds" table.
# - High-cardinality tags forbidden globally: user_id, request_id, raw_url, raw_query,
# raw_header_value, ip_address. These MUST NOT appear in any row.
# - tenant_id label is bounded mapping table id OR cohort bucket only (ULID raw forbidden).
# - error_code tag cardinality_limit follows error-codes.yaml row count (max 100).
metrics:
# === HTTP server (inbound) ===
# source: feature-metrics-alerting-contract — Metric/Alert Defaults
# "HTTP metric | http.server.requests with method/status/uri-template | raw URL or user id tag"
- name: http.server.requests
type: timer
unit: seconds
tags:
- name: method
cardinality_limit: 8
allowed_values: [GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS, OTHER]
- name: status
cardinality_limit: 7
allowed_values: [1xx, 2xx, 3xx, 4xx, 5xx, ok, other]
- name: uri_template
cardinality_limit: 200
validation: must_be_template_not_raw_uri
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p1: "error_rate > 5% for 5m OR > 10% for 1m"
p2: "error_rate > 1% for 10m"
p3: "error_rate > 0.1% for 1h"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: [method, status, uri_template]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — P1/P2/P3 정량 기준 (HTTP latency p99)
# "P1: p99 > 5s 5분 / P2: p99 > 1s 10분 / P3: p99 > 500ms 30분"
- name: http.server.requests.latency
type: timer
unit: seconds
tags:
- name: method
cardinality_limit: 8
allowed_values: [GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS, OTHER]
- name: uri_template
cardinality_limit: 200
validation: must_be_template_not_raw_uri
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p1: "p99 > 5s for 5m"
p2: "p99 > 1s for 10m"
p3: "p99 > 500ms for 30m"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: [method, uri_template, duration_ms]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === HTTP client (outbound dependency) ===
# source: feature-metrics-alerting-contract — Metric/Alert Defaults
# "dependency metric | dependency.client.requests with dependency.name/type/outcome | endpoint with secret tag"
# source: feature-outbound-http-client-baseline — "circuit breaker metric은 dependency.name, dependency.type, outcome까지만 tag로 허용"
- name: dependency.client.requests
type: timer
unit: seconds
tags:
- name: dependency_name
cardinality_limit: 50
- name: dependency_type
cardinality_limit: 10
allowed_values: [http, grpc, db, cache, queue, broker, other]
- name: outcome
cardinality_limit: 5
allowed_values: [SUCCESS, FAILURE, CIRCUIT_OPEN, TIMEOUT, REJECTED]
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p1: "required dep unavailable for 2m"
p2: "optional dep degraded for 5m"
p3: "spike alert (10x baseline)"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: [dependency_name, dependency_type, outcome, duration_ms]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-outbound-http-client-baseline — decisions
# "retry/circuit breaker 기본 라이브러리는 Resilience4j"
# source: feature-metrics-alerting-contract — "retry/CB minimum: resilience4j.retry.calls{outcome}"
- name: resilience4j.retry.calls
type: counter
unit: total
tags:
- name: name
cardinality_limit: 50
- name: outcome
cardinality_limit: 5
allowed_values: [SUCCESS, FAILURE, CIRCUIT_OPEN, TIMEOUT, REJECTED]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "retry exhaustion rate > 1% for 10m"
owner_branch: feature-outbound-http-client-baseline
log_field_mapping: [dependency_name, outcome, retry_attempt]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — "resilience4j.circuitbreaker.state"
- name: resilience4j.circuitbreaker.state
type: gauge
unit: total
tags:
- name: name
cardinality_limit: 50
- name: state
cardinality_limit: 6
allowed_values: [CLOSED, OPEN, HALF_OPEN, DISABLED, FORCED_OPEN, METRICS_ONLY]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p1: "state == OPEN for required dependency for 2m"
p2: "state == OPEN for optional dependency for 5m"
owner_branch: feature-outbound-http-client-baseline
log_field_mapping: [dependency_name]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — "resilience4j.circuitbreaker.calls{outcome}"
- name: resilience4j.circuitbreaker.calls
type: timer
unit: seconds
tags:
- name: name
cardinality_limit: 50
- name: outcome
cardinality_limit: 5
allowed_values: [SUCCESS, FAILURE, CIRCUIT_OPEN, TIMEOUT, REJECTED]
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p2: "CIRCUIT_OPEN rate > 1% for 10m"
owner_branch: feature-outbound-http-client-baseline
log_field_mapping: [dependency_name, outcome, duration_ms]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === DB connection pool ===
# source: feature-persistence-failure-baseline — Hikari Alert Threshold
# "pool wait p99 > 100ms 5분 지속 → P2 / pool exhaustion (active = max) > 1분 → P1"
# source: feature-metrics-alerting-contract — "hikaricp.connections.acquire{outcome='timeout'} p99 > 100ms"
- name: hikaricp.connections.acquire
type: timer
unit: seconds
tags:
- name: pool
cardinality_limit: 5
- name: outcome
cardinality_limit: 3
allowed_values: [SUCCESS, TIMEOUT, FAILURE]
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p1: "pool exhaustion (active == max) for 1m"
p2: "acquire p99 > 100ms for 5m"
owner_branch: feature-persistence-failure-baseline
log_field_mapping: [pool, outcome]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-persistence-failure-baseline — In scope "Hikari metric 노출 기준" + Hikari Alert Threshold
- name: hikaricp.connections.usage
type: timer
unit: seconds
tags:
- name: pool
cardinality_limit: 5
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p2: "usage p99 elevated > 10m"
owner_branch: feature-persistence-failure-baseline
log_field_mapping: [pool]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-persistence-failure-baseline — In scope "Hikari metric 노출 기준"
- name: hikaricp.connections.active
type: gauge
unit: total
tags:
- name: pool
cardinality_limit: 5
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p1: "active == max for 1m"
owner_branch: feature-persistence-failure-baseline
log_field_mapping: [pool]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — Histogram Buckets/Percentile "DB query: same"
- name: db.query.duration
type: timer
unit: seconds
tags:
- name: operation
cardinality_limit: 20
allowed_values: [select, insert, update, delete, batch, ddl, other]
- name: outcome
cardinality_limit: 3
allowed_values: [SUCCESS, FAILURE, TIMEOUT]
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p2: "p99 > 1s for 10m"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: [operation, outcome, duration_ms]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Background job / async executor ===
# source: feature-background-job-async-contract — Decisionized Work Items "saturation policy"
# "AbortPolicy default (core=10, max=50, queue=200)"
- name: executor.saturation
type: gauge
unit: total
tags:
- name: executor_name
cardinality_limit: 10
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "queue size > 80% capacity for 5m"
p1: "rejection rate > 0 for 1m"
owner_branch: feature-background-job-async-contract
log_field_mapping: [executor_name]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-background-job-async-contract — Decisionized Work Items "saturation | bounded executor + rejection log"
- name: executor.rejected.total
type: counter
unit: total
tags:
- name: executor_name
cardinality_limit: 10
- name: policy
cardinality_limit: 3
allowed_values: [AbortPolicy, CallerRunsPolicy, DiscardPolicy]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p1: "rejection_count > 0 for 1m"
owner_branch: feature-background-job-async-contract
log_field_mapping: [executor_name, policy]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-background-job-async-contract — Decisionized Work Items "retry/DLQ | exp backoff jitter, max 3, DLQ exhausted"
- name: job.retry.total
type: counter
unit: total
tags:
- name: job_name
cardinality_limit: 50
- name: outcome
cardinality_limit: 4
allowed_values: [SUCCESS, RETRY, EXHAUSTED, DLQ]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "EXHAUSTED rate > 1% for 10m"
owner_branch: feature-background-job-async-contract
log_field_mapping: [job_name, outcome, retry_attempt]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-background-job-async-contract — "DLQ after exhausted attempts"
- name: job.dlq.total
type: counter
unit: total
tags:
- name: job_name
cardinality_limit: 50
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p1: "DLQ rate sustained > 0 for 5m"
owner_branch: feature-background-job-async-contract
log_field_mapping: [job_name]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Distributed lock ===
# source: feature-distributed-lock-contract D7 — "metric lock.acquisition (tag: outcome =
# acquired/timeout/error) — 신규 제안" / D5 — try-lock + 유한 waitTime + lease(TTL). 분산
# 상호배제(distributedLockProvider) 획득 시도 결과를 센다. key 는 tag 로 넣지 않는다
# (무한 cardinality — 위 전역 금지 규칙). timeout outcome 은 LOCK_ACQUISITION_TIMEOUT 발생과 1:1.
- name: lock.acquisition
type: counter
unit: total
tags:
- name: outcome
cardinality_limit: 3
allowed_values: [acquired, timeout, error]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "timeout rate > 5% for 10m"
owner_branch: feature-distributed-lock-contract
log_field_mapping: [outcome]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-distributed-lock-contract §Edge / D5 (SI-LOCK-C5) — "lease 만료 후 unlock →
# ConcurrentModificationException — 삼킴 금지, 로그+metric 후 정상 흐름 복귀". Counts releases that
# found the lease already expired (the JdbcLock row was reclaimed by another instance before
# the holder called close()). A sustained nonzero rate means lease TTL is shorter than real
# critical-section duration — raise APP/lease TTL or shorten the protected work. Not an
# acquisition outcome, hence a separate counter from lock.acquisition.
- name: lock.lease.expired
type: counter
unit: total
tags: []
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "lease-expired rate sustained > 0 for 10m"
owner_branch: feature-distributed-lock-contract
log_field_mapping: []
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Outbox publisher ===
# source: feature-domain-event-outbox-contract — Outbox Defaults
# "DB outbox table with eventId, aggregateId, eventType, payload, occurredAt, status, attemptCount, nextAttemptAt"
- name: outbox.publisher.published.total
type: counter
unit: total
tags:
- name: event_type
cardinality_limit: 50
- name: outcome
cardinality_limit: 5
# The relay cycle outcomes (OutboxRelayResult.Outcome), not row statuses: IN_FLIGHT is a status
# and is never counted. STALE: the relay's claim was taken over before its result could be
# written (V17 lease fence). RELEASED: a stopping relay gave back a claim it had not used.
allowed_values: [PUBLISHED, FAILED, DEAD, STALE, RELEASED]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "FAILED rate > 1% for 10m, or STALE > 0 for 10m (in-flight timeout shorter than a relay stall)"
owner_branch: feature-domain-event-outbox-contract
log_field_mapping: [event_type, outcome, event_id]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-domain-event-outbox-contract — row status enum PENDING/IN_FLIGHT/PUBLISHED/FAILED/DEAD
- name: outbox.publisher.lag
type: gauge
unit: seconds
tags:
- name: event_type
cardinality_limit: 50
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "lag > 60s for 10m"
p1: "lag > 300s for 5m"
owner_branch: feature-domain-event-outbox-contract
log_field_mapping: [event_type]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-domain-event-outbox-contract — row status enum + Outbox Defaults attemptCount
# PARKED: an operator held a row that was retried without end (re-review 2026-11-03 RR-05-03).
- name: outbox.pending.size
type: gauge
unit: total
tags:
- name: status
cardinality_limit: 6
allowed_values: [PENDING, IN_FLIGHT, PUBLISHED, FAILED, DEAD, PARKED]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "PENDING size growing for 10m"
owner_branch: feature-domain-event-outbox-contract
log_field_mapping: [status]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: re-review 2026-11-03 RR-05-03 — a refusal for now and an unanswered publish are retried
# without end; these count the FAILED rows deferred at least 10 times (OutboxMetrics
# LONG_RETRY_DEFERRED_CLAIMS, the publish-failed runbook's §4.1 query) and age the oldest of them.
- name: outbox.retry.long.size
type: gauge
unit: total
tags: []
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "> 0 for 10m (runbook://outbox/publish-failed §4.1)"
owner_branch: feature-domain-event-outbox-contract
log_field_mapping: []
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
- name: outbox.retry.long.age
type: gauge
unit: seconds
tags: []
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "> 3600 (runbook://outbox/publish-failed §4.1)"
owner_branch: feature-domain-event-outbox-contract
log_field_mapping: []
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Cache ===
# source: feature-cache-consistency-contract — Decisionized Work Items "cache pattern | cache-aside default"
- name: cache.gets.total
type: counter
unit: total
tags:
- name: cache_name
cardinality_limit: 50
- name: result
cardinality_limit: 3
allowed_values: [hit, miss, error]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p3: "hit_ratio < baseline 0.5x for 1h"
owner_branch: feature-cache-consistency-contract
log_field_mapping: [cache_name, result]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-cache-consistency-contract — "invalidation = after-commit only", "invalidation 실패가 조용히 무시되면 실패"
- name: cache.invalidations.total
type: counter
unit: total
tags:
- name: cache_name
cardinality_limit: 50
- name: outcome
cardinality_limit: 3
allowed_values: [SUCCESS, FAILURE, SKIPPED]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "FAILURE rate > 0 for 5m"
owner_branch: feature-cache-consistency-contract
log_field_mapping: [cache_name, outcome]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: redis-production-capability — optional bounded cache-only L1
- name: cache.local.requests.total
type: counter
unit: total
tags:
- name: cache_name
cardinality_limit: 50
- name: result
cardinality_limit: 4
allowed_values: [hit, miss, error, bypass]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p3: "bypass or error rate above baseline for 15m"
owner_branch: redis-production-capability
log_field_mapping: [cache_name, result]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: redis-production-capability — local stale-age bound
- name: cache.local.entry.age.seconds
type: timer
unit: seconds
tags:
- name: cache_name
cardinality_limit: 50
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p3: "p99 approaches configured local TTL for 30m"
owner_branch: redis-production-capability
log_field_mapping: [cache_name]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: redis-production-capability — bounded invalidation and generation reconciliation
- name: cache.local.maintenance.total
type: counter
unit: total
tags:
- name: cache_name
cardinality_limit: 50
- name: event
cardinality_limit: 14
allowed_values:
- evict_cardinality
- evict_weight
- evict_ttl
- evict_invalidation
- flush_invalidation
- reconcile_generation_changed
- reconcile_unchanged
- reconcile_error
- subscriber_disconnected
- subscriber_overflow
- subscriber_malformed
- subscriber_publish_success
- subscriber_publish_error
- other
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "reconcile_error, subscriber_overflow, or sustained disconnects for 5m"
owner_branch: redis-production-capability
log_field_mapping: [cache_name, event]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Redis connection lanes ===
# source: review 2026-11-03 RR-04-02 — the cache's conditional writes got their own lane
# (CACHE_SCRIPT) so that cache fills cannot take the SCRIPT lane's connections from idempotency,
# rate limiting and leases. These meters show, per lane, whether that holds. Emitted by
# redis-runtime RedisLaneMetrics for every lane RedisRuntimeOwner pools, only when the context has
# a MeterRegistry. Every counter series is registered at zero at startup.
- name: redis.lane.acquire.wait.seconds
type: timer
unit: seconds
tags:
- name: lane
# The lane's name: REGULAR, SCRIPT (redis-runtime), CACHE_SCRIPT (cache-redis), BLOCKING,
# TRANSACTION, ADMIN (redis-sdk), plus any lane a service composes as a RedisLaneContribution.
cardinality_limit: 12
- name: outcome
cardinality_limit: 2
allowed_values: [admitted, refused]
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p2: "SCRIPT admitted p99 > 100ms for 10m (the correctness lane is queueing)"
p3: "CACHE_SCRIPT admitted p99 > 100ms for 30m"
owner_branch: redis-production-capability
log_field_mapping: [lane, outcome]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
- name: redis.lane.acquire.refused.total
type: counter
unit: total
tags:
- name: lane
# The lane's name: REGULAR, SCRIPT (redis-runtime), CACHE_SCRIPT (cache-redis), BLOCKING,
# TRANSACTION, ADMIN (redis-sdk), plus any lane a service composes as a RedisLaneContribution.
cardinality_limit: 12
- name: reason
cardinality_limit: 4
allowed_values: [saturated, closing, interrupted, connect_failed]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p1: "SCRIPT or REGULAR saturated or connect_failed above zero for 2m while a correctness role is selected"
p2: "CACHE_SCRIPT saturated above baseline for 10m"
owner_branch: redis-production-capability
log_field_mapping: [lane, reason]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
- name: redis.lane.connections.opened.total
type: counter
unit: total
tags:
- name: lane
# The lane's name: REGULAR, SCRIPT (redis-runtime), CACHE_SCRIPT (cache-redis), BLOCKING,
# TRANSACTION, ADMIN (redis-sdk), plus any lane a service composes as a RedisLaneContribution.
cardinality_limit: 12
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p3: "opened rate after warm-up above baseline for 15m (reconnect churn against a slow Redis)"
owner_branch: redis-production-capability
log_field_mapping: [lane]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
- name: redis.lane.connections.discarded.total
type: counter
unit: total
tags:
- name: lane
# The lane's name: REGULAR, SCRIPT (redis-runtime), CACHE_SCRIPT (cache-redis), BLOCKING,
# TRANSACTION, ADMIN (redis-sdk), plus any lane a service composes as a RedisLaneContribution.
cardinality_limit: 12
- name: reason
cardinality_limit: 2
allowed_values: [invalidated, broken]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p3: "invalidated rate above baseline for 15m (timed-out or failed calls discarding connections)"
owner_branch: redis-production-capability
log_field_mapping: [lane, reason]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
- name: redis.lane.connections.active
type: gauge
unit: total
tags:
- name: lane
# The lane's name: REGULAR, SCRIPT (redis-runtime), CACHE_SCRIPT (cache-redis), BLOCKING,
# TRANSACTION, ADMIN (redis-sdk), plus any lane a service composes as a RedisLaneContribution.
cardinality_limit: 12
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "SCRIPT active == configured maximum-script-connections for 5m"
owner_branch: redis-production-capability
log_field_mapping: [lane]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
- name: redis.lane.connections.pending
type: gauge
unit: total
tags:
- name: lane
# The lane's name: REGULAR, SCRIPT (redis-runtime), CACHE_SCRIPT (cache-redis), BLOCKING,
# TRANSACTION, ADMIN (redis-sdk), plus any lane a service composes as a RedisLaneContribution.
cardinality_limit: 12
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "SCRIPT pending > 0 for 5m"
owner_branch: redis-production-capability
log_field_mapping: [lane]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Log appender ===
# source: feature-log-management-contract — Sampling Policy (final)
# "async appender overflow default: drop oldest INFO/DEBUG with counter metric (log.appender.dropped.total)"
- name: log.appender.dropped.total
type: counter
unit: total
tags:
- name: appender
cardinality_limit: 5
- name: level
cardinality_limit: 2
allowed_values: [INFO, DEBUG]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "dropped > 0 sustained for 10m"
owner_branch: feature-log-management-contract
log_field_mapping: [appender, level]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === Distributed tracing ===
# source: feature-distributed-tracing-contract — decisions
# "trace sampling rate default = prod 1%, dev/local 100%; shipped runtime profiles are local|dev|prod"
- name: tracing.sampling.rate
type: gauge
unit: total
tags:
- name: profile
cardinality_limit: 3
allowed_values: [prod, dev, local]
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p3: "effective rate deviates from configured for 1h"
owner_branch: feature-distributed-tracing-contract
log_field_mapping: [profile]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# === JVM baseline ===
# source: feature-metrics-alerting-contract — In scope "JVM/process metric"
- name: jvm.memory.used
type: gauge
unit: bytes
tags:
- name: area
cardinality_limit: 2
allowed_values: [heap, nonheap]
- name: id
cardinality_limit: 10
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p2: "heap used / max > 0.85 for 10m"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: [area, id]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — In scope "JVM/process metric"
- name: jvm.gc.pause
type: timer
unit: seconds
tags:
- name: action
cardinality_limit: 10
- name: cause
cardinality_limit: 10
percentiles: [0.5, 0.9, 0.95, 0.99]
histogram_buckets: slo_driven
alert_severity_thresholds:
p2: "p99 pause > 500ms for 10m"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: [action, cause]
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — In scope "JVM/process metric"
- name: jvm.threads.live
type: gauge
unit: total
tags: []
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p3: "thread count > 2x baseline for 30m"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: []
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality
# source: feature-metrics-alerting-contract — In scope "JVM/process metric" (process uptime)
- name: process.uptime
type: gauge
unit: seconds
tags: []
percentiles: null
histogram_buckets: null
alert_severity_thresholds:
p1: "uptime reset unexpectedly < 60s (crash loop signal)"
owner_branch: feature-metrics-alerting-contract
log_field_mapping: []
compatibility_impact: additive
required_test: contract-verification:metrics-cardinality