Checkout has been stable for 15 minutes. A postmortem follows.
Incident management
Run incidents without switching tools
Declare an incident from the dashboard, from Slack or straight from an alert. Assign a commander, post updates, keep customers informed on your status page and write the postmortem in one place.
- A Slack channel per incident
- Opened, announced and kept in sync.
- Roles and severities
- Commander, communications lead, SEV1 to SEV4.
- Postmortems built in
- Started from the incident's own timeline.
Timeline
One timeline from declared to resolved
Each update moves the incident from investigating to resolved. Roles, severity changes, updates and status page posts land on one timeline, whether they happened in the dashboard or in Slack.
- SEV1 to SEV4
- Updates in Markdown
- Every change with the person who made it
- Kestrel
- Incidents
- INC-12
Checkout API returning 503 errorsResolvedSEV1Critical
Status for the next updateSelect a stage to move the incident forward
Timeline
- Sat 24 OctUTC
Published to Kestrel status: resolved
ResolvedLena Visser MonitoringLena Visser The rollback is out and checkouts go through again. We're watching error rates.
The alert resolved: 200 OK from Europe
IdentifiedSam Okafor The last deploy broke the payment provider's callback. We're rolling it back.
Sam Okafor became communications lead
Published to Kestrel status: Checkout, major outage
InvestigatingLena Visser Checkout requests are failing for some customers. We're looking into it.
Created #inc-12-checkout-api in Slack
Lena Visser became commander
Linked the alert Checkout API is down
Lena Visser declared the incident from the alert
Slack
Declare, update and resolve from Slack
Type /incident declare, use the shortcut on a message or press the button on an alert. Spectate opens a channel such as #inc-42-checkout-errors and keeps it in sync with the dashboard.
- /incident declare, update, roles, status-page, resolve
- Pinned header that stays in sync
- Actions attributed to the person's account
SpectateAPP13:07
INC-12 · Checkout API returning 503 errors
Severity SEV1Status InvestigatingCommander Lena VisserStarted 13:07
Post updateStatus pageRolesResolveView in Spectate
SO
Sam Okafor13:12
Identified: the last deploy broke the payment callback. Rolling it back.
Posted with /incident update · published to Kestrel status
Communication
Tell customers from the same screen
Publish to your status page from the incident, in the dashboard or from Slack: pick the affected components, write the update and choose whether to email subscribers.
- Impact per component
- Optional email to subscribers
- Kestrel
- Status pages
- Kestrel status
Postmortem
A postmortem that starts half written
Start from a template that already holds the facts and a timeline built from every update, then write the rest in Markdown.
- Pre-filled facts and timeline
- Root cause and action items
- Kestrel
- Incidents
- INC-12
- Postmortem
Checkout API returning 503 errorsDraft
Blameless write-up of what happened, why, and what changes as a result.
postmortem.md
58 words · 9 lines
Markdown
## Summary
A deploy renamed the payment callback variable, so checkouts failed for 10 minutes.
## Timeline
Imported from the incident: 9 events, 4 updates.
## Follow-up
- [x] Check required env vars at boot
- [ ] Canary before the full rollout
Incident
- Status
- Resolved
- Severity
- Critical
- Duration
- 25m 12s
- Updates
- 4
In detail
Everything in incident management
Alerts are not incidents
Alerts are signals; incidents are the human response.
Incident numbers
INC-42, the same in the dashboard and in Slack.
Announcements channel
New incidents announced in a Slack channel you choose.
Automatic incidents
An escalation step can open one, with its Slack channel.
Slack user linking
Slack users act with their own Spectate permissions.
Backdated updates
Set when it happened, from the dashboard or Slack.
Included on every plan. See pricing
FAQ
Incident management: questions
How do I declare an incident in Slack?
Install the Spectate Slack app from Integrations, then type /incident declare followed by an optional title, or use the Declare incident shortcut on any message. A modal asks for the title, severity and summary. Spectate creates the incident, opens a dedicated channel if that setting is on, and posts an announcement.
Does every incident get its own Slack channel?
Yes, by default. Whether an incident is declared in Slack, in the dashboard or by an escalation path, Spectate creates a channel such as #inc-42-checkout-errors with a pinned header and actions. You can turn this off and change the channel prefix in the Slack settings.
What severities and stages does Spectate use?
Severities run from SEV1 (critical) to SEV4 (low). Incidents move through four stages: investigating, identified, monitoring and resolved. Each update sets the stage, and severity changes are recorded on the timeline with the person who made them.
Can alerts open incidents automatically?
Yes. Add a create incident step to an escalation path and Spectate opens an incident when that step runs, with a Slack channel if Slack is connected. Slack alert messages also have a Declare incident button for alerts that need a person to decide.
Does Spectate help write postmortems?
Yes. Each incident has a postmortem editor that starts from a template with the severity, start and resolution times, duration and a timeline built from the incident's updates. You add the summary, impact, root cause, contributing factors, resolution and action items in Markdown, then mark it published.
Can I update my status page from an incident?
Yes. From the incident in the dashboard, or with /incident status-page in Slack, choose the status page, the affected components and their impact, write the message and decide whether to email subscribers. The public page's component status and uptime bars follow the impact you set.
Know about it first.
Set up your first monitor in under a minute. 14 days free, no credit card.