Module 72: Production Operations
Uptime targets
🏠The everyday picture
A bus company promises "buses every 10 minutes, 95% of the time". It's a clear promise, it can be measured, and it accepts that a few late buses happen.
📘What it really is
-
SLI (service level indicator): something you measure, such as the share of requests that succeed, or the share served within 1 second.
-
SLO (service level objective): your internal target for it, such as "99.5% of requests succeed each month".
-
SLA (service level agreement): a promise to customers in a contract, usually looser than the SLO, often with refunds if broken.
-
Error budget: the failures your SLO allows. At 99.5% per month, about 3.6 hours of downtime (or the equivalent in errors). While budget remains, ship features; when it's used up, focus on reliability first.
-
Choose targets by need (Module 67): an internal CRM used in office hours may be fine at 99.5%; a payments service needs far more. Each extra nine costs more (Module 68).
-
Planned maintenance still counts against users; schedule it for quiet hours and announce it.
- A simple status page (many uptime services include one) tells users what's happening during problems, which saves everyone time.
🖱️Try it
Write SLOs for your CRM: an availability target and a speed target, with how each is measured and the monthly error budget in minutes. Check your uptime service's report after a week and compare.
✅Quick check
What's an error budget?
Show the answer
The amount of failure an SLO allows; while it lasts you can take risks, when it's spent you focus on reliability.
What's the difference between an SLO and an SLA?
Show the answer
An SLO is an internal target; an SLA is a promise to customers, often with penalties.