Skip to content
KaDeep Docsv4.0
Open StudioDownload

Synthetic monitors

For managers and QA leads2 min readUpdated for v4.0

Your view

A synthetic monitor watches your app between releases. It opens a page, or replays one of your tests, on a schedule. If it fails, KaDeep opens an incident and tells you.

1. Open Synthetic monitors

Section titled “1. Open Synthetic monitors”
In the sidebar, open Workspace › Synthetic Lab. Press Ctrl K (Windows) or ⌘ K (Mac), type Synthetic monitors, and press Enter.
Synthetic monitors: the summary, New monitor, the Incidents tab, and a monitor with its state, uptime and latency (Manage view)
Synthetic monitors: the summary, New monitor, the Incidents tab, and a monitor with its state, uptime and latency (Manage view)

Manage view

Synthetic monitors: the summary, New monitor, the Incidents tab, and a monitor with its state, uptime and latency (Hands-on view)
Synthetic monitors: the summary, New monitor, the Incidents tab, and a monitor with its state, uptime and latency (Hands-on view)

Hands-on view

  • 1 How many monitors are up, degraded, down and paused, and how many incidents are open.
  • 2 New monitor adds one.
  • 3 Monitors and Incidents.
  • 4 Each monitor: what it checks and how often, its state, the project, uptime in the last 24 hours, latency (p50 and p95), its last checks and how it alerts.
State Meaning
Up The last check passed in time.
Degraded It passed, but slower than your limit.
Down The last check failed. An incident is open.
Paused It doesn’t run until someone resumes it.

Availability 30 d (CSV) and Availability 90 d (CSV) download uptime for a service report.

2. Add a monitor

Section titled “2. Add a monitor”
  1. Choose New monitor.
  2. What to check: choose Page check and type the address, or Replay a test and choose one of the project’s tests.
  3. Schedule: how often it checks, in minutes.
  4. Pass when: what the page must return, and the time limit. A slow check counts as degraded.
  5. Alerts: in the app, on Slack or by email, and after how many failed checks. Maintenance mute silences it for a few hours.
  6. On-call: who is paged, and Page again every so many minutes until someone acknowledges.
  7. Maintenance: weekly windows, for example every Saturday at 02:00 for 2 hours. Checks in a window don’t page anyone and don’t count against uptime.
  8. Save the monitor.

3. Handle an incident

Section titled “3. Handle an incident”

Open the Incidents tab. Each incident shows the monitor, when it started, and the failed checks. Open a check to see the run with its steps and screenshots.

  • Acknowledge tells the team you are on it, and stops repeat pages.
  • Resolve closes it once the monitor is back up.

To stop a monitor for a while, open it and choose Pause. Resume starts it again.

What’s next

Section titled “What’s next”
Monitoring and reliabilityLive operations and flaky tests.
Schedule runsRun whole suites nightly or after each deploy.
IntegrationsConnect Slack for alerts.