Skip to content

Reference

Limits

This article collects the defaults and limits stated in the rest of the documentation. Where a value is a default, it can be changed; a dash means the documentation states no setting.

Retention and windows#

ItemValueConfigurable
Raw telemetry retention90 daysAgreed per deployment
Aggregate retentionLonger than rawAgreed per deployment
Historical log window60 minutes—
Usage history in the consoleToday and the last 30 days—

Detection#

ItemValueConfigurable
Baseline historyPrevious seven days, same time of dayDefault
Source silenceFive times the expected intervalDefault
Recovery of a numeric alertTwo consecutive valid evaluationsConfigured threshold
Probe or health check failureTwo consecutive failed runsDefault
Probe or health check recoveryTwo consecutive successful runsDefault
Observed edge inactivity24 hoursYes
Cause candidatesWithin three hops of the symptom—

See Baselines, Source health, Synthetic checks, Service map and Cause assessment.

Sources and allowances#

ItemValue
Allowance alertsAt 80 % and 100 % of the daily allowance
Default cap modeSoft: requests are accepted and overage is flagged
Hard capRequests receive 429 with Retry-After
Active credentials per sourceTwo
Metering dayThe tenant time zone

Requests and lists#

ItemValue
List page size50 by default, 200 at most
Request sizeLimited; over the limit receives 413
Request rateLimited; over the limit receives 429 with Retry-After
Default retry interval after 429 or 503One second

Size and rate limit values are set per deployment. Limits are transport errors: telemetry is never silently sampled or dropped to stay within them (see OTLP endpoint).

Delivery#

ItemValue
TicketsOne per qualifying incident
Forwarding-failure alertTwo consecutive minutes of failures, major
Queue capacityConfigured per destination

Roles and approvals#

ItemValue
Roles per person in a tenantOne
ApprovalSingle level; the requester cannot approve their own request
Action write accessOff until granted

Last updated October 2026