Incidents · Plans and data
Limits
Plans differ in how many incidents a month they include. Past the limit nothing stops: an outage doesn't wait for a new month.
Limits by plan
Loading the limits…
What counts
Each incident opened in the project counts once, in the calendar month it was opened (in UTC), whether a person declared it or a call through the API opened it. The count starts again on the 1st.
- A call that joins an open incident by its dedupe key isn't a new one.
- An incident later declined still counts: it was opened.
- Updates, status changes, notices, follow-ups and write-ups don't count.
Past the limit
Every incident is still opened, counted and shown as over: declaring one is never refused, because one can start in the middle of an outage. Nothing is taken away either.
You see where the month stands:
- at the foot of Incidents' sidebar, how many of how many, on every page;
- on Incidents' Settings, with what the plan allows;
- by email: the project's admins hear at 80 % and when the month's are used up, once each a month;
- on the workspace's Usage page, for its owners, with the last six months.
A bigger plan is a change on the workspace's Subscription page at id, charged at once for the rest of the period. See Incidents' page for the plans.
Other limits
The same on every plan:
| What | Limit |
|---|---|
| API calls | 120 a minute per server secret |
| Title | 200 characters |
| Summary and each update | 5,000 characters |
| Each part of the write-up | 10,000 characters |
| A follow-up | 200 characters |
| When it began | Up to 30 days back, never later than now |
| On the status page | While open, and for 14 days once resolved |
When the subscription ends
Your incidents stay readable, but nothing new is taken in or changed: the pages are read only, and the API answers 402. Renewing at id opens it all again.