About this interactive
A service level agreement (SLA) sets the level of service a provider promises and what happens when it falls short. Uptime is the usual measure.
Allowed downtime. Take the time in the period and multiply it by the part not promised. A 30-day month has 30 × 24 = 720 hours, or 43,200 minutes. At 99.9% uptime, 0.1% may be down: 43,200 × 0.001 = 43.2 minutes a month. Over a 365-day year (8,760 hours) the same 99.9% allows 8.76 hours, the lesson's "about 8 hours a year".
Measured uptime. Subtract the downtime from the period and divide by the period: 216 minutes down in a 30-day month gives (43,200 − 216) ÷ 43,200 = 99.5%.
Service credits. The SLA's table sets bands. The questions here use one table: 99.9% or more, no credit; 99.0% up to 99.9%, 10% of the month's bill; 95.0% up to 99.0%, 25%; under 95.0%, 100%. Find the band the month's uptime falls in, then take that share of the bill. A band's lower edge belongs to it, so exactly 99.0% earns 10%.
The lesson reads Amazon S3's SLA as giving 10% between 98% and 99.9%. In S3's published SLA for its standard storage class, the 10% band runs from 99.0% to 99.9%; the 98% edge belongs to the storage classes that promise 99%. That is why the table here uses 99.0%.
An SLA you hold from a vendor gives you recourse if the service falls short. One you give your own customers sets their expectations, and gives them the same recourse against you.
About TechKnowSurge
TechKnowSurge builds IT and cybersecurity professionals through hands-on, concept-first training built around real understanding — not memorization. Free interactive tools, structured programs, and 25+ years of real-world experience, all in one place.
Explore free tools and programs →