Service Level Agreement

SUS Service Level Agreement

Effective: Whenever we last remembered to update this

Binding In: All known dimensions, plus three we're still mapping

Our Commitment

We guarantee 99.9% uptime for all production services.*

Yes, 99.9%. Not 99.99%. We prefer to under-promise and over-deliver.

Our actual uptime is 99.999%. This is suspicious.

*We decide what counts as "production" whenever we feel like it. Hint: if you're paying us, it's production.

What Counts as Downtime

• Services are unreachable in this dimension
• Services are unreachable in adjacent dimensions (parallel universe coverage included)
• Services return errors for your valid requests
• Services return correct data for invalid requests (this has happened once, we're still investigating)
• Services are significantly degraded (>50% slower than normal, or time is moving backwards)
• The Hydrocoptic Marzlevane is experiencing side fumbling

What Doesn't Count as Downtime

• Scheduled maintenance (we'll tell you 72 hours ahead, in your local timezone, unless you're in Antarctica)
• Your code having bugs (we checked, it's definitely your code)
• Internet issues between you and us (we blame Comcast preemptively)
• Force majeure (acts of nature, war, penguin uprisings, etc.)
• Quantum Flux Capacitor recalibration (takes exactly 47 seconds, every time, suspiciously consistently)
• Services working too well (this actually triggers our anomaly detection)
• Scheduled Prefabulated Amulite rotations (quarterly, unless Mercury is in retrograde)
• That one time Sterling accidentally unplugged something

The Suspicious Fine Print

Your SLA coverage begins the moment you sign up and extends into perpetuity, or until the heat death of the universe, whichever comes first. In the event of universal heat death, credits will be prorated.

SLA calculations are performed by our proprietary Uptime Verification Engine™, which has never been wrong. This is suspicious. We audit the auditor quarterly.

Credits

If we miss our SLA:

Credits are applied automatically. You don't have to:

• File a ticket
• Talk to someone
• Prove you were affected
• Sacrifice a goat
• Navigate a 47-page claims process

We just... give you the money back. Suspicious, we know.

How We Calculate Uptime

Uptime = (Total Minutes - Downtime Minutes) / Total Minutes × 100

Where:
- Total Minutes = All the minutes (we count them)
- Downtime Minutes = Minutes things weren't working (we count these too)
- Math = Correct (we checked)

We round in your favor. Always. Our accountant finds this suspicious.

Multi-Dimensional Availability Zones

Our services are replicated across:

• 7 physical regions (on this planet)
• 3 parallel dimensional planes (for redundancy)
• 1 pocket dimension (for cold storage)
• The cloud (which is just someone else's computer, but nicer)

In the event of a dimensional collapse, services will failover to the nearest stable reality. You may notice brief inconsistencies in the space-time continuum. This is normal.

Reporting

We publish real-time status at suspicious.cloud/status.

We publish monthly uptime reports. They're suspiciously positive.

We don't hide incidents. Even the embarrassing ones:

• The Great Penguin Interference of 2024
• The time we achieved 101% uptime and had to file an incident for being too available
• That one Tuesday when everything was fine but everyone kept checking because it was suspicious

Termination Clause

You can leave anytime. We'll be sad, but we'll export your data within 24 hours in any format you want.

If we decide to terminate services, we'll give you:

• 90 days notice (minimum)
• Help migrating (we'll even be nice about it)
• A refund for any unused credits
• A formal apology from Sterling

We've never had to do this. Suspicious.

Questions about our SLA? [email protected]

Want to test our SLA? Please don't. But also, go ahead. We're curious now.