Troubleshooting

No data in DevTools

  • Confirm Uyava.initialize() runs before the app starts.
  • Run the app in debug or profile mode (VM Service must be available).
  • Ensure the Uyava extension is selected in DevTools.
  • Verify your graph snapshot uses unique IDs.

Desktop cannot connect

  • Paste the VM Service URI exactly as shown in your terminal.
  • Confirm the app is running with VM Service enabled (debug/profile).
  • If you are behind a firewall or VM, ensure the URI is reachable.

No Uyava output in app console

  • Confirm Uyava.enableConsoleLogging(...) is called in app startup.
  • Check minLevel: if too high, lower-severity events are intentionally hidden.
  • Check includeTypes / excludeTypes: you may be filtering out all events.
  • If you disabled logging earlier with Uyava.disableConsoleLogging(), re-enable it.
  • For full config options, see SDK Integration.

Filters hide everything

  • Clear the filters and check the graph again.
  • Invalid regex or mask patterns emit diagnostics and are ignored.
  • Remember that focus only affects the Journal, not the graph.

Diagnostics warnings

  • nodes.duplicate_id means the last payload wins.
  • edges.dangling_* means an edge referenced a missing node and was dropped.
  • nodes.invalid_color or nodes.invalid_shape indicates invalid styling.

Use the Diagnostics Docs button to see the exact fix.

.uyava logs are missing

  • For SDK logging, ensure directoryPath is valid and writable.
  • On macOS sandboxed builds, logs may be stored under the container temp folder.
  • Use Uyava.exportCurrentArchive() to force a sealed file for sharing.

.uyava file does not open in Desktop

  • Make sure you are testing with the latest Desktop build.
  • Open the file from Desktop File mode (not only by double-click association).
  • Check archive integrity:
gunzip -c session.uyava > /tmp/session.ndjson
  • If decompression works but Desktop still fails to load the file, report the issue with platform + Desktop version.

Performance tips

  • Collapse parent groups to reduce visual noise.
  • Avoid a single synthetic root parent for all nodes; split into meaningful top-level roots to reduce graph crowding.
  • Use severity filters for incident triage.
  • Enable sampling in file logging to reduce volume.