The services we measure, over the last day.
98.7% of 8,372 samples from UtiBox server (single origin) found services up. One location; regional problems elsewhere won’t show here.
- 1Outlook101 of 624 failed
- 2MongoDB Atlas18 of 563 failed
- 3Microsoft 3653 of 623 failed
- 4Figma2 of 563 failed
- 5VRChat2 of 563 failed
- 6Jira & Confluence1 of 563 failed
- 7Microsoft Azure1 of 563 failed
- 8Netflix1 of 624 failed
- 9Telegram1 of 624 failed
- 10LinkedIn1 of 624 failed
How to read the pulse
Every service in the catalogue is requested from our server on a schedule. Failed samples per hour counts the samples that found a service slow, erroring or unreachable, and the line shows the average response time that hour, so a busy hour and a broken hour look different.
User reports per hour counts the problems people reported on status pages. Reports are shown on their own: a service’s state comes from our measurements, not from reports.
Most failed samples ranks services by failed samples over the last 30 days. It needs at least a day of samples for a service before it can rank it.
All of this is measured from one place, UtiBox server (single origin). A failure here can be a problem between our server and the service rather than everywhere, which is why each status page also shows checks from other regions when they’re available.