Alerts & notifications FAQ
What gets alerted, where it goes, how to silence.
What does Sikurd alert on?
Out of the box:
- Instance Down — two consecutive missed polls.
- Trunk Down — a SIP trunk fails to register.
- License Threshold — extension usage crosses 80% / 95%.
- License Expiry — 30 / 7 / 1 days before expiry.
- Backup Failed — backup ran but errored.
- Backup Stale — no successful backup in 7+ days.
- High Disk — > 90% disk used.
- Calls Threshold — concurrent calls cross 80% / 95% of licensed capacity.
- Service Down — a 3CX-side service fails.
Where do alerts go?
Per the routing matrix you configure. Options:
- Mobile push (free)
- Email (free)
- Slack / Teams / Webhook (Pro+)
- PSA ticket (Pro+)
- AI voice call (Pro+, Dialora-powered)
Why didn't I get an alert I expected?
Most common causes:
- Email verification pending — polling pauses until you verify.
- No routing configured for that type/severity — check the matrix.
- Quiet hours — alerts during your tenant's quiet window are suppressed for non-critical channels.
- Trial ended — polling paused.
- Maintenance window — alerts are suppressed for the instance during a planned window.
Can I snooze an alert without resolving?
Yes — click Mute for N hours on the alert row. Suppresses any re-fires of the same alert for the duration.
Can multiple people get the same alert?
Yes. Configure recipients per channel in Alert routing → Recipients. Mobile push goes to everyone with the app installed; email to everyone on the recipient list.
Mobile app
iOS + Android apps available. Install from app stores, then sign in with your Sikurd credentials. Push notifications require granting permission on first sign-in.
Voice calls
AI voice calls use Dialora. The agent reads "Alert: [type] on [instance]" — terse so the recipient knows what's wrong but no internal details leak. Configure recipients + agent voice in Settings → Integrations → Dialora.
Snoozing whole instances
Use Maintenance windows on the instance detail page to suppress all alerts for a scheduled downtime.
False positives
If you're getting alerts that don't reflect real problems, click the … menu on the alert row → Adjust thresholds. Tweaks the trigger for that alert type, scoped to the instance or fleet-wide.