Troubleshooting
Delayed Or Sparse Data
Use this guide when Vight shows data, but it looks delayed, intermittent, or lighter than expected.
Symptom
Use this guide when Vight shows data, but it looks delayed, intermittent, or lighter than expected.
Recent requests appear after a delay or only in some views. Counts look lower than the traffic you expected. A chart or list looks empty until filters are widened.
Likely causes
- Small sample
- Low traffic services can look sparse even when instrumentation is correct.
- Window mismatch
- The chart, list, and detail page may be filtered to a window that misses the event you are checking.
- Different label
- Data can appear delayed or absent when it landed under another service, environment, version, or deployment.
Checks
Start with the broadest possible evidence window, then narrow the cause once you know where data is landing.
- 1
Widen the time range and remove service, environment, severity, status, and text filters.
- 2
Generate a fresh request and a fresh log event, then wait briefly before refreshing.
- 3
Compare Services, Traces, Logs, and Runtime to see which signal is sparse.
- 4
Check whether traffic is going to a different environment, version, deployment, or service name.
Resolution
The fix should make new telemetry appear in the expected Vight page, under the expected service and environment, without relying on stale data.
Use wider filters while validating setup, then narrow the investigation. Align service identity values across deployments. Use traces and logs together when one view has fewer examples. Open the signal-specific troubleshooting guide if only one data type is missing.
After applying the fix, validate it this way: Start with a broad time range and remove optional filters. Generate a fresh request and log event, then refresh after a short wait. Check Services, Traces, Logs, and Runtime to find which signal is sparse. Once visible, narrow filters one at a time until the empty state returns.