Troubleshooting
Troubleshooting common issues
Reconnects, limits, stuck sends, and when to contact support.
Last updated: May 2026
Quick triage
| Symptom | First checks |
|---|---|
| Nothing sends | Campaign active? Inside send window? LinkedIn connected + recent sync? |
| Sudden stop mid-day | Likely weekly/daily invite limit or provider 422 cannot_resend_yet—wait or lower limits |
| Errors mention auth / 401 | Reconnect LinkedIn hosted auth; then sync |
| Duplicate invite blocked | 24h duplicate guard or true duplicate—expected safety |
| Sequence not progressing | Prospect replied (automation cancelled)? Campaign inactive? Worker backlog? |
| Message stuck “sending” | Wait & refresh; >60s UI may mark failed—verify thread for double-send before retrying |
Error prefix legend (when shown)
UNI:— Upstream provider (Unipile / LinkedIn semantics). Often includes HTTP code + subtype text.APP:— Fulreach internal classification (DB, queue, disconnected account, duplicate prevention).
Capture screenshots + timestamps + affected prospect public URL when opening support—accelerates cross-checking send logs.
Healthy operating practices
- Pilot on narrow ICP samples.
- Avoid editing windows aggressively during live peak send minutes (some reschedule paths still maturing across dual backend stacks).
- Keep 1 primary strategic owner per LinkedIn identity to avoid cross-tool invite collisions outside Fulreach guard windows.
Escalation
Persistent 502 waves, checkpoint loops, or billing/invite activation mismatches need human review with workspace id, account id, campaign id (copy from URL or admin exports).