[ SRE_ARENA ]
Roadmap
System Design
SREdle
⚡ 1420 ELO
OP
☰
←
BACK TO ROADMAP
SLOS & ERROR BUDGETS
100% Uptime is the Wrong Goal.
Visualization: 1. The Mathematics of Nines
Two Nines
99.0%
7h 12m
Allowed Downtime / mo
Three Nines
99.9%
43.8 min
Allowed Downtime / mo
Five Nines
99.999%
26 sec
Allowed Downtime / mo
Press enter or space to select a node. You can then use the arrow keys to move the node around. Press delete to remove it and escape to cancel.
Press enter or space to select an edge. You can then press delete to remove it or escape to cancel.
1. The Mathematics of Nines
2. Error Budgets: The Ultimate Negotiator
3. Burn Rate Alerting
DOWNLOADING CONCEPT: 1. THE MATHEMATICS OF NINES...
↳
Business leaders always ask for 100% reliability.
↳
Your job as an SRE is to explain why that's mathematically impossible and financially ruinous.
↳
Let's do the math for a 30-day window: -
99% (Two Nines):
7 hours, 12 minutes of downtime allowed.
↳
-
99.9% (Three Nines):
43.8 minutes of downtime allowed.
↳
-
99.99% (Four Nines):
4.38 minutes of downtime allowed.
↳
-
99.999% (Five Nines):
26 seconds of downtime allowed.
↳
Going from 99.9% to 99.99% costs exponentially more money and engineering effort (you need active-active multi-region failovers, etc.).
↳
Does the business actually lose money if the system is down for 40 minutes a month? If not, a 99.99% SLO is a waste of engineering resources.
↳
Set the SLO exactly where the user starts to feel pain, and no higher.
▲ LIVE INCIDENT
The CEO wants to sign an SLA guaranteeing 99.999% uptime to a new enterprise customer.
sre@arena:~/slos-error-budgets
sre@arena
:
~
$
uptime_calculator --nines 5 --window 1_year
▶ EXECUTE COMMAND