Four products. One place to look.
Uptime, incidents, logs, and status pages share one account, one bill, and one incident timeline. Each one is good enough to replace the tool you use for it today.
-
Uptime monitoring
HTTP, keyword, ping, port, DNS, SSL, and heartbeat checks every 30 seconds from twelve regions. A failure is confirmed from a second region before anyone is woken up.
- 30-second checks from 12 regions, confirmed before paging
- Screenshot, response, traceroute, and MTR on every failure
- SSL, domain, and heartbeat monitors included
- Response
- Screenshot
- Timeline
Captured 03:14:02 · Frankfurt, DE · 1440×900
-
Incident management and on-call
Rotations, overrides, and a three-step escalation ladder. Pages land in Slack, Teams, SMS, or a phone call, and acknowledging from any of them stops the rest.
- Follow-the-sun schedules with one-tap overrides
- Similar incidents merge into one so nobody is paged ten times
- Post-incident timeline written for you, minute by minute
On-call this week Platform · follow the sun-
Priya Natarajan
Primary · until Thu 09:00
-
Tomás Ferreira
Secondary · escalates after 5 min
-
Engineering manager
Escalates after 15 min
-
Log management
Every line is ingested, indexed, and kept. Query with SQL, PromQL, or drag and drop, and drop the noise at the edge before it is billed.
- No sampling: ad-hoc queries over raw logs at any volume
- Drop rules stop lines being ingested and being billed
- Anomaly alerts on the shape of a metric, not a threshold
Query Sampling off2,567,345 rows in 0.7 s SQL · PromQL · Drag and drop -
Status pages
A branded page on your own subdomain that your customers subscribe to. Incidents post to it automatically, and the uptime bars are drawn from the same checks that paged you.
- Custom domain, colours, logo, and CSS on every plan
- Per-component email and RSS subscriptions
- Uptime and p95 response time published from real checks
Northwind Status
All systems operational
How it holds up
The questions engineers ask before they move their on-call to us.
Where do the checks run from?
Twelve regions across North America, South America, Europe, Asia, Africa, and Australia. Business plans can add private probes inside your own network for internal services.
How do you avoid false alarms?
A failure has to be seen from a second region before anyone is paged, and the interval between the two is under ten seconds. Flapping transit hops show up in the traceroute, not in your phone.
Can Carbon replace our incident tool?
For most teams, yes. Schedules, overrides, escalation ladders, phone and SMS paging, and post-incident timelines are all included. Existing schedules import from a calendar feed.
Does the status page update itself?
When a monitor fails, the incident it opens posts to the status page automatically, and the resolution posts when the monitor recovers. You can edit the wording at any point, or turn automatic posting off per component.
Sleep through the next one.
Set up your first monitor in two minutes. The free plan pages you, and stays free.
Start monitoring for free or book a demo
- Virginia 21 ms
- Oregon 58 ms
- São Paulo 112 ms
- Dublin 74 ms
- Frankfurt 82 ms
- Stockholm 91 ms
- Mumbai 188 ms
- Singapore 204 ms
- Tokyo 156 ms
- Sydney 221 ms
- Cape Town 240 ms
- Toronto 29 ms