> ## Documentation Index > Fetch the complete documentation index at: https://docs.generalvalidation.com/llms.txt > Use this file to discover all available pages before exploring further. # Troubleshooting and logs > Diagnose setup, dataset preparation, validation, delivery, and evidence access issues across the hosted app and your Azure execution environment. Start with the affected page in General Validation. Record the typed error code, request ID when shown, and the relevant Run, Test, or operation ID. The hosted application reports metadata diagnostics; execution details and exact data evidence stay in your Azure environment. ## Find the right surface | Surface | Use it for | | -------------------------- | ----------------------------------------------------------------------------------- | | Environment | Customer installation, provisioning, repair, and readiness blockers | | Discovery and Datasets | Catalog access, import, schema preparation, and dataset readiness | | Results | Run progress, accepted verdicts, per-Test errors, and execution identifiers | | Browser result preview | Direct access to exact metrics and captured rows in your Storage account | | Customer ADF monitoring | Pipeline, activity, source permission, and execution failures | | [Error reference](/errors) | Stable error-code meanings and recommended remedies | | General Validation support | Hosted API or background-processing failures, correlated by request or operation ID | ## Route the failure | Symptom | What to check next | | ------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------- | | Sign-in or membership denied | Confirm the signed-in Microsoft account, Organization directory connection, active membership, and required role. | | Run or Pair creation blocked by entitlement | Open **Billing** and review trial/subscription status and Pair capacity. Ask an Owner to resolve billing actions. | | Environment not ready | Open **Environment**, inspect the reported resource steps, and use the available provisioning or repair flow. | | Dataset not ready | Open the Dataset, review its reason and remediation, correct the source issue, then use the offered preparation action. | | Submission rejected | Read the UI message or API problem response. This rejection occurs before a Run exists. | | Execution failed | Open the affected Run and Test, then use its execution identifier in customer ADF monitoring. | | Validation failed | Review the saved rule, thresholds, reported counts, and available customer-side evidence. The comparison completed but did not meet its rule. | | Result delivery failed after ADF succeeded | Use **Retry incomplete items** when offered to resume delivery on the same Run. If it remains blocked, share the typed error and IDs with support. | | `result_callback.delivery_mismatch` | Contact support with the Run and submission IDs. The delivered result summaries do not match this Run; retrying delivery will not resolve the conflict. | | Reconnecting or status checks paused | Use **Retry status check** to resume status reads. It does not create a new Run. | | Concurrent edit rejected | Reload the current resource, review its changes, and then retry the edit. API clients must obtain the current ETag after a `412`. | Some temporary ADF infrastructure failures receive one automatic retry. Follow the existing Run's progress before submitting the same work again. If a Run has both delivery and execution failures, **Retry incomplete items** first recovers delivery. After that finishes, retry any remaining eligible execution failures. ## Resolve a catalog or Fabric access issue Use **Reload catalog** or **Retry refresh** after a failed browser read to load saved catalog metadata again. Use **Retry catalog refresh** when the Azure catalog refresh itself failed. Previously loaded rows can remain visible while a refresh is unsuccessful. For Fabric sources, check the specific failure: | What you see | What to check | | ----------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Workspace is missing from the setup picker | The Organization Owner connecting it must also be a Fabric workspace Admin or Member. Check the selected Microsoft account and any requested consent. | | Workspace access needs repair | Open **Manage Fabric access** and use **Verify Viewer access** or **Repair Viewer access** as offered. Review the customer execution identity shown there. | | `discovery.fabric_capacity_inactive` | Ask a Fabric capacity administrator to check the workspace's capacity assignment and resume or activate it, then retry. Changing Viewer permissions does not resolve inactive capacity. | | A lakehouse or warehouse's table listing failed | Review that item's error and retry its table listing when offered. Check its SQL endpoint readiness, customer execution access, and network route. Previously listed tables may remain visible but unavailable for a new import until listing succeeds. | See [Connect Microsoft Fabric](/connect-microsoft-fabric) for the access requested during setup. ## Resolve source access and schema issues Catalog access and execution access are separate. A discoverable object can still lack a supported schema, accepted storage route, or permissions for the customer execution identity. Check the customer ADF identity's access to the exact Storage path or SQL/Fabric table and its network route. Correct source settings in your own environment, then use **Refresh metadata**, **Retry preparation**, or **Recheck Dataset** as offered. See [Supported scope](/supported-scope) for accepted formats and schemas. Grant source-data permissions to the customer execution identity identified by the setup flow. The hosted application's identity does not need customer Storage data access to validate or render results. ## Resolve a missing or denied result preview The browser reads artifacts directly from your registered Storage account using your signed-in identity. If the verdict appears but exact values or rows do not, check: * Result delivery has completed and the selected Test captured the evidence you expect. * Your Microsoft account has the required delegated Storage consent and data permissions. * Your browser can reach the registered customer Storage endpoint through your organization's network rules. * The artifact still exists, matches the registered account, and has not been removed by a retention policy. After resolving the issue, use the preview's **Retry** control. A zero failed-row capture limit means no row evidence was requested; rerunning with capture enabled is a separate validation decision. See [Runs, results & evidence](/runs-results-and-evidence). ## Logs and support correlation General Validation operates the hosted API and background-service logs. Use the application diagnostics and your customer ADF monitoring for initial investigation. Support can correlate a hosted failure using the request ID shown by the UI or returned in `x-request-id`, plus the Run or operation identity. The browser also writes structured application diagnostics to its local Developer Tools console. These records stay local; browser diagnostics are not automatically sent to the hosted service. Review any record before sharing it. When contacting support, include the action you attempted, approximate UTC time, error code, request ID, and Run or operation ID. Keep exact values, failed rows, credentials, tokens, and raw customer response bodies in your own environment. Do not attach an unreviewed network archive or evidence screenshot that includes customer data. Use your customer ADF monitoring and Azure diagnostics for execution history. Set diagnostic and Storage retention according to your organization's needs.