Target Service Level Objective (SLO)
"You're allowed 8h 45m of panic per year."
Per Year
8h 45m 36s
Total allowed downtime annually.
Per Month
43m 48s
Allowed monthly outage window.
Per Week
10m 04s
Weekly maintenance budget.
Per Day
1m 26s
Daily incident threshold.
Reality Check (What you can do while the server burns)
Enough time to fly from London to New York.
~8 hours and 45 minutes. You could technically board a plane, cross the Atlantic, and land before your customers realize the site is back up.
Watch LOTR: Fellowship of the Ring (Extended Edition) twice.
With a runtime of 228 minutes, you can fit in two full viewings. That's a lot of walking to Mordor while your DevOps team walks through hell.
Complete Elden Ring if you're really good.
Speedrunners finish it in under 30 minutes, but for mortals, ~8 hours might get you past the first boss. Git gud while your servers are down.
Slow-cook a perfect brisket from start to finish.
A proper Texas-style brisket takes 8-12 hours. Your services are down, but at least dinner will be legendary.
Cost of a Second
At 99.9% availability, you could lose up to $50,000 annually from downtime.
NASA aims for 7 nines (99.99999%) for the space shuttle software.
You are not NASA. Your cat walking on the keyboard will cost you more than your budget allows.
Detailed Breakdown Table
| AVAILABILITY | DOWNTIME / YEAR | DOWNTIME / MONTH | DOWNTIME / WEEK | REAL WORLD ANALOGY |
|---|