Your notification channel is down
A messenger, email gateway or integration fails. The alert fires, but nobody receives it. You may only find out from the consequences.
Use case
A backup alert channel
An additional route to the engineer on call. Connect Acktide to your monitoring alongside existing notifications, then test it when the primary channel is unavailable.
When the primary channel fails
The message arrived. Nobody responded.
A notification in a shared chat does not assign responsibility. You need an explicit ACK from a named person, and escalation if nobody acknowledges it.
One alert, two channels
Alertmanager sends the critical event to both receivers at the same time. There is nothing to switch during an incident; the backup is always active for the rules you choose.
| Channel | What the team receives |
|---|---|
| Existing channel | Your usual notifications in a messenger, email or current on-call system. Existing settings stay in place when you add a parallel path. |
| Acktide | An Android system notification on the lock screen, ACK directly from the notification, and escalation to a backup engineer if nobody acknowledges it. |
Existing Slack, email and webhook receivers in Alertmanager keep working unchanged; you add one more receiver.
What makes a backup independent
Host the backup outside your primary site, and check that monitoring and phones can reach Acktide if that site fails.
- Separate hostingAcktide’s server and database run outside the cluster or site whose failure you are protecting against.
- Phone access during a primary network outageThe phone must be able to reach Acktide when the primary network is unavailable. Include VPN and DNS in your test plan.
- An independent signal sourceIf an entire site goes down, the alert must originate elsewhere: external checks or separate monitoring.
- Tested recoveryTest restoring from a backup and administrator access in advance.
What the backup can and cannot do
| Situation | Acktide behavior |
|---|---|
| Primary notification channel fails | Independent delivery to the phone while the alert source, Acktide and the network remain available |
| Phone temporarily offline | Delivery resumes after reconnection if the server has already received and stored the event |
| Entire site unavailable | Requires Acktide outside that site and an independent signal source; see site peering |
| No phones reachable | Delivery waits for connectivity. SMS and calls require a separately configured provider and are disabled by default |