The right alert to the right person
Escalation paths decide who hears about what: notify a channel, page whoever is on call or the whole team, wait, branch on severity or working hours. Nobody acknowledges? Page again, page the next person, or start over.
- 9 channel types
- From email and Slack to SMS and phone calls.
- Branches and working hours
- Split a path on severity, time of day, source or labels.
- Your other tools too
- Grafana, Alertmanager, Sentry, webhooks and email.
Escalation paths you can read at a glance
Notify a channel, page whoever is on call, the next person or chosen members, wait, open an incident. Branch on severity, working hours, source, labels or on whether anyone has acknowledged yet.
- Wait a fixed time or until working hours begin
- Still not acknowledged after 30 minutes? Branch on it
- Page again, or repeat the path up to 10 times
- Group an outage into one page, by path, project, resource or label
- Maintenance windows, once or every week, that mute alerts and keep uptime clean
- Multiple paths on Team and Business, one on Starter
- Kestrel
- Alerting
- Database escalation
- Ifwithin working hours
- orseverity is critical
Alerts where your team already is
Email, Slack, Microsoft Teams, Discord, Telegram, Pushover, webhooks, SMS and phone calls. Messages update in place when an alert is acknowledged or resolved.
- Acknowledge from Slack, Telegram or a phone call
- Say "not me" or snooze from Slack and Discord; "not me" from Telegram and a call too
- Page someone by hand, from an alert, an incident or Slack
- Response time and metric charts in the alert
- SMS and voice credits included every month
- Kestrel
- Alerting
Destinations your paths notify. Send a test to check delivery.
One place for alerts from your other tools
Connect Grafana, Prometheus Alertmanager, Sentry, any webhook or plain email. Repeats keep one alert firing, and a resolve from the tool resolves it.
- Duplicates collapse into one firing alert
- Resolves when the tool says it's resolved
- Last 7 days of raw requests for debugging
- Kestrel
- Alerting
- Sources
- Grafana
Recent events
Everything this source received in the last 7 days.
- firedHigh error rate on checkout-api2m ago
- resolvedPostgresReplicationLag on db-fra-248m ago
- still firingHigh error rate on checkout-api1h ago
- firedDisk will fill in 4h on fra-db-013h ago
- testTest alert sent from Spectate1d ago
Everything in alerting
Alert mails that explain
The cause in plain words, and what to check first.
Failing channels switch off
Escalations skip them and owners get an email.
Test any channel
A successful test re-enables a disabled channel.
Severity mapping
P1, fatal or warning from your tools, mapped for you.
Escalation log
Every step and branch decision on the alert's timeline.
Paths are snapshotted
Edits never disturb a running escalation.
Included on every plan. See pricing
Alerting: questions
Which notification channels does Spectate support?
Email, Slack, Microsoft Teams, Discord, Telegram, Pushover, webhooks, SMS and phone calls. Every channel type is available on every plan. SMS and phone calls use credits: every plan includes a monthly allowance of 10, 50 or 200 credits, and packs of 100 cost €19 and never expire.
What is an escalation path?
A set of steps that runs when an alert fires: notify channels, wait, page people, open an incident. Branches send alerts down different sides based on severity, working hours, the source, labels or whether someone acknowledged the alert. Paging stops as soon as someone acknowledges, and steps that can't deliver are skipped.
What happens when nobody acknowledges an alert?
That is up to the path. A paging step can page the same people again every few minutes, the next step can page whoever is on call next, everyone on the schedule or the whole team, and a repeat step goes back to an earlier step up to 10 times. A branch can also wait, for example 30 minutes, and check whether the alert is still not acknowledged.
Can I alert differently during and outside working hours?
Yes. Each escalation path has working hours with days, times and a time zone. A branch can check whether an alert arrives within or outside them, and a wait step can hold an alert until working hours begin, for example to page at night only for critical alerts and wait until morning for the rest.
Can Spectate receive alerts from Grafana, Prometheus or Sentry?
Yes. Add an alert source for Grafana, Prometheus Alertmanager, Sentry, a generic webhook or email. Spectate turns each request into an alert with a severity, routes it through your escalation paths and resolves it when the tool reports recovery. Repeated notifications for the same problem keep one alert firing.
How do I acknowledge an alert?
From the dashboard, with the Acknowledge button in Slack, Discord or Telegram, or by pressing 1 when Spectate calls you. Acknowledging stops the paging, and the messages update to show who acknowledged it and when. Any team member can acknowledge.
What happens if a notification channel stops working?
Spectate counts failed deliveries per channel. After 5 failures in a row, 2 permanent errors such as a revoked token, or 24 hours of failures, the channel is disabled and owners and admins get an email. Escalations skip it and continue with the next step, and a successful test re-enables it.
Do I need the Team plan for escalation paths?
Starter includes one escalation path, the team default, with every channel type. Team and Business add as many paths as you need and steps that page people: whoever is on call, chosen members or the whole team. Team costs €0.60 per monitor a month with a minimum of 80 monitors.
Know about it first.
Set up your first monitor in under a minute. 14 days free, no credit card.