Link Exceptions to Operations
Use this query to correlate exceptions to request operation IDs for end-to-end failure tracing.
Data Source
| Table | Schema Note |
exceptions | App Insights table. Requires telemetry pipeline configured for the app. |
Query Pipeline
flowchart LR
A[Filter by cloud role name] --> B[Project exception fields] --> C[Join mentally with requests on operation_Id] --> D[Sort by time]
Query
let AppName = "my-container-app";
exceptions
| where cloud_RoleName == AppName
| project timestamp, type, outerMessage, operation_Id
| order by timestamp desc
Example Output
| timestamp | type | outerMessage | operation_Id |
| 2026-04-04T11:45:14.219Z | RuntimeError | upstream dependency timeout | 9f7a7d9d0bb84f0b |
| 2026-04-04T11:45:12.917Z | PermissionError | token acquired but storage access denied (403) | f0e5946f613c4a49 |
| 2026-04-04T11:45:10.600Z | ConnectionError | connection refused by backend service | 4dbfc25be8c74999 |
Interpretation Notes
- Match
operation_Id with failed requests to find root exception per user call. - Recurring
type often indicates a single dominant fault class. - Normal pattern: infrequent exceptions and rapid recovery.
Limitations
- Exception telemetry volume can be sampling-limited.
- Does not include infrastructure-only failures without app telemetry.
See Also