The authoritative view is Web App → App Service Plan → Quotas; if every chart now reads zero, Azure does not expose a separate documented F1 “quota-stop history” log that reliably names the previously breached counter, so use App Service diagnostics and Activity Log to capture the platform event before escalating.
Identify the enforced quota
- Open the affected web app in the Azure portal.
- Select App Service Plan → Quotas.
- Expand and record every chart’s:
- quota name
- current limit
- current value
- Resets in value
For Free and Shared plans, the relevant enforcement behavior is: ^1^
| Quota |
Enforcement and reset |
| CPU (Short) |
Stops the app; resets every five minutes |
| CPU (Day) |
Stops the app; resets at midnight UTC |
| Bandwidth/Data Out |
Stops the app; resets at midnight UTC |
| Memory |
Stops the app temporarily |
| Filesystem |
Writes fail, including log writes; it does not match the documented quota-stop 403 behavior |
Exceeding CPU Short, CPU Day, or Bandwidth causes App Service to stop the app until reset, with incoming requests returning HTTP 403. The portal’s Resets in value is the documented indicator of the remaining enforcement period. ^2^
A zero current value does not identify which counter caused an earlier stop. A closely matching F1 case reports that the displayed meter can reset to zero while the app remains in the exceeded/stopped state from the preceding measurement window. ^3^
Because the corrected startup command concerns the earlier 503, application startup logs are not the primary evidence for this current platform-generated 403.
- Go to Web App → Diagnose and solve problems.
- Open Availability and Performance.
- Run:
- App Down Workflow
- Web App Down
- CPU Usage
- Memory Usage
- Linux - Number of Running Containers
- Set the examined interval to include the first transition to Quota exceeded.
- Save screenshots or export the report, including timestamps and any detector finding.
App Service diagnostics requires no prior logging configuration and is most useful for events from the previous 24 hours. ^4^
Then open Web App → Activity log and filter around the same time for:
- Restart Web App
- Stop Web App
- other administrative or resource-health events
Activity Log is the monitoring surface for web-app resource operations; application logs and Log Stream are intended primarily for code/runtime output. ^5^
Do not expect Python application logs to contain the quota rejection: requests blocked by the App Service front end do not reach the worker, and front-end failures are not recorded as back-end HTTP server errors. ^1^
If the app remains stopped after the reset
If CPU (Short) remains blocked beyond five minutes, or CPU (Day)/Bandwidth remains blocked after midnight UTC, capture:
- subscription and resource identifiers
- App Service plan name, F1 SKU, Linux, and West US 3
- time the state first changed to
QuotaExceeded
- quota-chart screenshots showing zero usage, limits, and reset values
- Activity Log events and correlation IDs
- App Service diagnostic report
- the successful Start operation followed by the immediate return to
QuotaExceeded
Open Help + support → Create a support request and select Web app down or reporting errors under Function or Web App (Windows and Linux). Ask support to identify the internal quota dimension and clear or explain the stale enforcement state. A regional SKU quota-increase request applies to failed plan creation or scaling, not to an already-created F1 app stopped by runtime CPU, bandwidth, or memory enforcement.
If immediate recovery is required, scale the App Service plan above Free/Shared; Microsoft documents upgrading as the way to increase or remove these runtime quotas. ^1^
References
- Azure App Service Quotas and Metrics - Azure App Service
- Error 403 - This web app is stopped message - Azure
- Web app free plan - Quota exceeded even tho quotas shows 0 usage - Microsoft Q&A
- Diagnostics in Azure App Service
- Monitor Azure App Service