specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Azure Monitor
providerId: microsoft-azure-monitor
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-05'
reconciled: true
tags:
  - Rate Limiting
  - Observability
  - Microsoft Azure
description: Azure Monitor surfaces multiple APIs each with separate rate limits - Log Ingestion API
  (Data Collection Rules), Logs Query API, Metrics API, Application Insights track API. Daily ingestion
  cap can be set per workspace. Query throttling is enforced per workspace and per user.
sources:
  - https://learn.microsoft.com/en-us/azure/azure-monitor/service-limits
  - https://learn.microsoft.com/en-us/azure/azure-monitor/logs/api/timeouts
  - https://learn.microsoft.com/en-us/azure/azure-monitor/logs/daily-cap
headers:
  retryAfter: Retry-After
responseCodes:
  throttled: 429
  serviceUnavailable: 503
limits:
  - name: Log Ingestion API per DCR throughput
    scope: data_collection_rule
    metric: bandwidth
    limit: 'see Azure Monitor service limits (default a few MB/sec per DCR)'
  - name: Logs Query API queries per user per workspace
    scope: user/workspace
    metric: queries_per_30_seconds
    limit: 200
    timeFrame: second
  - name: Logs Query API concurrent queries per user
    scope: user
    metric: concurrent_queries
    limit: 5
  - name: Logs Query API maximum data scanned per query
    scope: query
    metric: data_scanned
    limit: '64 GB per query (Analytics tier default)'
  - name: Logs Query API query timeout
    scope: query
    metric: timeout_seconds
    limit: 600
    timeFrame: second
  - name: Application Insights track API per ingestion
    scope: app_insights_resource
    metric: events_per_second
    limit: 'see App Insights ingestion sampling and rate caps'
  - name: Daily ingestion cap (configurable)
    scope: workspace
    metric: gb_per_day
    limit: 'workspace-configurable; set to bound daily spend'
  - name: Metrics API requests per subscription
    scope: subscription
    metric: requests_per_minute
    limit: 'see Azure Monitor REST API limits'
policies:
  - name: Honor Retry-After
    description: When ingestion or query APIs return 429 / 503, wait the duration in Retry-After
      and retry with exponential backoff.
  - name: Use sampling
    description: For Application Insights at high event rates, enable adaptive sampling to control
      cost and rate without code changes.
  - name: Daily cap to bound spend
    description: Set workspace daily cap to prevent runaway ingestion; cap triggers an alert when
      hit and stops ingestion until reset.
  - name: Use the Logs Ingestion API (DCR-based) for high volume
    description: New collectors should use the DCR-based Logs Ingestion API rather than the legacy
      Data Collector API for better throughput and schema control.
  - name: Batch and pre-aggregate metrics
    description: Pre-aggregate high-cardinality custom metrics client-side and emit per minute rather
      than per request to control sample-count costs.
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com