Reading the status page
What the daily bars, the components and the incident updates on the status page mean, and why some hours show as no data.
3 min read
The daily barsCopy link to this section
The status page shows the last 90 days as 90 bars, oldest on the left. Each bar is one day in Japan time. The workspace checks itself every five minutes, and a bar's colour is the share of that day's checks that found it up. Point at a bar to see the day and its percentage.
- Green: at least 99% of the day's checks were up.
- Amber: from 90% to just under 99%.
- Red: under 90%.
- Grey: no data. No check ran that day.
The figure under the bars is the uptime over the 90 days, counted over the checks that ran. Until there are any, it says that data is being collected.
Why some time shows as no dataCopy link to this section
The checks run inside the workspace itself. While it is asleep, rebuilding or restarting, nothing is checked, so that time is left out instead of being counted as up or as down. A day with a few missing hours keeps its colour from the checks that did run; a day with none is grey.
Note
No data is not the same as down. If people could not reach the workspace during that time, the incident list below the bars is where it is written up.
ComponentsCopy link to this section
Below the bars, each part of the service is listed as the workspace sees itself right now: Site, Sign-in, App, Database and Backups. Each is Operational, Degraded or Down. Only names and states are shown; nothing about how the workspace is set up.
The dot in every page's footer sums these up in one phrase: All systems operational, Some parts degraded or Service disruption.
Incidents and their updatesCopy link to this section
An incident is written up by whoever runs the workspace. Each one has an impact (Minor, Major or Maintenance), when it started and ended in Japan time, or ongoing while it is open, and its updates, oldest first:
| Update | Meaning |
|---|---|
| Investigating | The problem is known and being looked into. |
| Identified | The cause is found and a fix is under way. |
| Monitoring | A fix is in place and being watched. |
| Resolved | The incident is over. |
Each incident's title opens a page of its own with how long it lasted, the parts it affected and every update, so you can link to it.
Follow it from elsewhereCopy link to this section
- Feed: subscribe to incidents and their updates in any feed reader, from the link under the incident list.
- Badge and panel: a small image for a README or an intranet page, and a panel another site can show. Both are at most a minute old, and both are in Japanese with
?lang=ja.
Last updated