Azure Container Apps logs never reach Log Analytics through either supported logging path

We have an Azure Container Apps managed environment in Central US where live log streaming works, but no console or system logs reach Log Analytics through either supported configuration path.

Observed behavior

• Container app live streams show fresh stdout/stderr from two separate apps.
• One always-warm app emitted a controlled, unauthenticated server error at 2026-10-05 16:58:34 UTC and again at 17:03:46 UTC.
• A fresh, same-image zero-traffic revision emitted startup logs at 17:05:50 UTC and became Healthy with one replica.
• A second app was restarted and returned HTTP 200 from its health endpoint.
• Despite this, all relevant Log Analytics tables remained empty.

Classic workspace integration tested

• Managed environment provisioning state: Succeeded.
• appLogsConfiguration.destination: log-analytics.
• Workspace is in the same region, public ingestion/query are enabled, no daily cap, and the environment has no VNet integration.
• Microsoft.App, Microsoft.Insights, and Microsoft.OperationalInsights providers are registered.
• ContainerAppConsoleLogs_CL and ContainerAppSystemLogs_CL returned zero rows for all apps over 30 days.
• We re-applied the same workspace configuration using the current primary shared key without exposing or storing it.
• We restarted an app, created new same-image revisions, and toggled the destination to none and back to the same workspace. The tables remained empty.

Azure Monitor diagnostic-settings path tested

• Changed the environment destination to azure-monitor; provisioning reached Succeeded.
• Created an environment-level diagnostic setting to the same workspace with ContainerAppConsoleLogs and ContainerAppSystemLogs enabled.
• ARM recorded both diagnostic-setting writes as Succeeded / HTTP 200. Correlation IDs:
  • 51546c3f-7182-4f48-a08e-b8f14f1c6f94
  • c16b5750-4c90-4a5d-b0b1-4a10d3a8f6ef
• Azure provisioned ContainerAppConsoleLogs and ContainerAppSystemLogs as Analytics tables.
• ContainerAppConsoleLogs, ContainerAppSystemLogs, and AzureDiagnostics still returned zero rows. The oldest new event was more than 13 minutes old, and the newly created revision was more than five minutes old.
• Raw ARM readback confirmed destination azure-monitor, environment Succeeded, the correct workspace resource, and both categories enabled.

We rolled the diagnostic test back cleanly: removed the diagnostic setting, restored log-analytics with the current workspace key, and deactivated the zero-traffic test revision. Production remained healthy and at 100% traffic throughout.

Question

Is this a known managed-environment logging pipeline failure? What supported action can reinitialize or repair logging for this existing Container Apps environment without recreating the environment? Are there additional platform-side diagnostics Microsoft can check using the correlation IDs above?

No secrets, tenant IDs, subscription IDs, workspace IDs, or full resource IDs are included here.