Interactive demo
Take an incident from alert to review.
Fire an alert, watch the escalation policy page the on-call, acknowledge from a phone, tell customers on the status page, resolve, and start the postmortem.
The demo mimics AckNexa’s behaviour as the product guide describes it, simplified: there is no server, pushes, SMS and email are only drawn on this page, and the demo runs escalation in seconds instead of minutes.
Then try the reports: incidents per service, time to acknowledge per team and pages per person.
Try this
- Fire the P1 alert from Icinga.Not done yet.
- Wait about ten seconds without acknowledging. Step 2 pages Sam T.Not done yet.
- Tap the push on the phone and acknowledge there.Not done yet.
- See that the public status page still shows Checkout as Operational, and select “Publish an update” in its note.Not done yet.
- Publish the update and watch Checkout change to Major outage.Not done yet.
- Resolve the incident, or send Recover from Icinga.Not done yet.
- Draft the postmortem.Not done yet.
Monitoring
Each button sends the event that tool would send. Repeats of a firing alert are deduplicated; an incident resolves when every alert in it has recovered.
- P1Icinga 2OK
Database primary unreachable
checkout-db-01 · Checkout
- P2PrometheusOK
Checkout p99 latency above 2 s
checkout-web · Checkout
- P3PrometheusOK
Disk 90% full on search-02
search-02 · routed to Search
Maintenance window
Checkout incidents created during a window are recorded but not paged.
Respond
Incidents
Public status page
A preview of “Example status”. A service changes only when a responder publishes an update; internal titles and incident numbers never appear.
Administer
Reports
Sample data for the demo, plus the incidents you create here.
In AckNexa, Reports is for organization and team admins, and team admins see only their own teams. This is the organization admin’s view of Example Operations: about 90 days of generated history for five services, four teams and seven people on weekly rotations.
Incidents per service
Which services page the most and how quickly they are handled
| Service | Incidents | P1/P2 | Ack | Resolve |
|---|
Time to acknowledge per team
Median, slowest tenth, and how often the team answered within 5 minutes
| Team | Incidents | Median ack | p90 ack | ≤ 5 min | Resolve |
|---|
Incidents per day
When incidents start
By severity
Pages per person per week
How the load is spread. Darker cells mean more pages that week; after-hours counts nights and weekends.
The demo’s setup
Read-only. In AckNexa, organization administrators configure all of this in the Web application.
Organization and teams
Example Operations
- Payments: Priya N. and Sam T. (responders), Alex M. (team admin). Schedules: Payments primary, with Priya N. on call now, and Payments secondary, with Sam T.
- Search: Jordan K. (responder, on call now on Search primary), Alex M. (team admin).
The sample history in Reports adds the Database team (Mei L. and Tom B., taking turns weekly) and the Identity team (Dana K., with Tom B. as team admin), and the services Payments API, Orders DB and Identity. In it, Priya N. and Sam T. swap primary and secondary each week.
Services and integrations
- Checkout: owned by Payments, escalation policy Payments on-call, listed on the status page as Checkout.
- Search: owned by Search, escalation policy Search on-call, listed on the status page as Product search.
- Icinga 2 and Prometheus Alertmanager integrations default to Checkout. A routing rule sends Prometheus alerts with
labels.team = searchto Search. - Source severity maps to P1 for critical, P2 for high and P3 for warning.
Escalation policy: Payments on-call
- Immediately: primary on-call (Priya N.) by push.
- After 10 minutes without acknowledgement: secondary on-call (Sam T.) by push and SMS.
- After 10 more minutes: team admins (Alex M.) by email.
Search on-call
- Immediately: primary on-call (Jordan K.) by push.
- After 10 minutes: team admins (Alex M.) by email.
The demo runs escalation in seconds: each policy minute takes one second here. Neither policy repeats.
Status page and maintenance
Example status is a public page listing Checkout and Product search. A P1 or P2 incident with a published update shows its service as Major outage; lower severities show Degraded performance.
The maintenance window covers Checkout for 30 minutes, 30 seconds in the demo.
What the demo simplifies
- There is no server, database or worker: this page plays every event itself and keeps the state in this tab.
- Pushes, SMS and email are only drawn. Every push lands on one phone, and nothing reaches a provider or a person.
- Each policy minute runs as one second, and the maintenance window lasts 30 seconds.
- Each incident holds one alert. AckNexa can group several alerts into one incident and resolves it when all of them have recovered.
- The closing status update after a resolve appears at once; in AckNexa a worker publishes it shortly afterwards.
- The reports use sample data: about 90 days of history for Example Operations, generated in this tab and the same for everyone on a given day, plus the incidents you create here. AckNexa builds its reports on the server from your real incidents and pages.
- Subscriber emails, stakeholders, tickets, chat channels and the rest of the product are left out.
How it works in the product: Routing and deduplication · Escalation policies · Respond to an incident · Status pages · Postmortems