- 필수 기능
- 시작하기
- Glossary
- 표준 속성
- Guides
- Agent
- 통합
- 개방형텔레메트리
- 개발자
- API
- Datadog Mobile App
- CoScreen
- Cloudcraft
- 앱 내
- 서비스 관리
- 인프라스트럭처
- 애플리케이션 성능
- APM
- Continuous Profiler
- 스팬 시각화
- 데이터 스트림 모니터링
- 데이터 작업 모니터링
- 디지털 경험
- 소프트웨어 제공
- 보안
- AI Observability
- 로그 관리
- 관리
Error Tracking provides fine-grained control of which errors to ingest, helping you reduce noise and avoid unexpected costs.
You can define what data is included in Error Tracking in two ways:
You can configure both rules and rate limits on the Error Tracking > Settings page.
Rules allow you to select which errors are ingested into Error Tracking. They apply to both billable and non-billable errors.
Each rule consists of:
service:my-web-store
.env:staging
query to exclude staging errors.A given rule can be toggled on or off. An error event is included if it matches a query in one of the active inclusion filters and it does not match any active nested exclusion queries.
Note: Error events that get accepted by a rule might still be excluded from Error Tracking if they lack the required attributes.
Rules are evaluated in order, with the evaluation stopping at the first matching rule. The priority of the rules and their nested filters depends on their order in the list.
By default, Error Tracking has an *
inclusion filter and no exclusion filters. This means all error with the requirements to be fingerprinted are ingested into Error Tracking.
To add a rule (inclusion filter):
Rate limits allow you to control the number of billable errors included in Error Tracking per day. This cap applies to all errors that match the filters of a rule.
After the daily cap is reached, ingestion stops until the next day. You can modify or remove the cap at any time.
To set a rate limit:
A Rate limit applied
event is generated when you reach the rate limit. See the Event Management documentation for details on viewing and using events.
You can monitor your Error Tracking on Logs usage by setting up monitors and alerts for the datadog.estimated_usage.error_tracking.logs.events
metric, which tracks the number of ingested error logs.
This metric is available by default at no additional cost, and its data is retained for 15 months.