The Nines Have It…
There are numerous cliches and buzzwords we hear daily that creep into our lexicon without warrant of origin or meaning.
One of them that you’re undoubtedly used to hearing relates to the measurement of availability expressed as a percentage: the dreaded “nines.”
I read a story this morning on the launch of the “Stratus Trusted Cloud” that promises the following:
Since it is built on the industry’s most robust, scalable, fully redundant architecture, Stratus delivers unmatched performance, availability and security with 99.99% SLAs.
It’s interesting to note what 99.99% availability means within the context of an SLA — “four nines” means you have the equivalent of 52.6 minutes of resource unavailability per year. That may sound perfectly wonderful and may even lead some to consider that this exceeds what many enterprises can deliver today (I’m interested in the veracity of these claims.) However, I would ask you to consider this point:
I don’t have access to the contract/SLA to know whether this metric refers to total availability that includes both planned and unplanned downtime or only planned downtime.
This is pretty important, especially in light of what we’ve seen with other large and well-established Cloud service providers who offer similar or better SLA’s (with or without real fiscal repercussion) and have experienced unplanned outages for hours on end.
Is four nines good enough for your most critical applications? Do you measure this today? Does it even matter?
/Hoff
Here’s a handy Wikipedia reference on availability table you can print out:
| Availability % | Downtime per year | Downtime per month* | Downtime per week |
|---|---|---|---|
| 90% | 36.5 days | 72 hours | 16.8 hours |
| 95% | 18.25 days | 36 hours | 8.4 hours |
| 98% | 7.30 days | 14.4 hours | 3.36 hours |
| 99% | 3.65 days | 7.20 hours | 1.68 hours |
| 99.5% | 1.83 days | 3.60 hours | 50.4 minutes |
| 99.8% | 17.52 hours | 86.23 minutes | 20.16 minutes |
| 99.9% (“three nines”) | 8.76 hours | 43.2 minutes | 10.1 minutes |
| 99.95% | 4.38 hours | 21.56 minutes | 5.04 minutes |
| 99.99% (“four nines”) | 52.6 minutes | 4.32 minutes | 1.01 minutes |
| 99.999% (“five nines”) | 5.26 minutes | 25.9 seconds | 6.05 seconds |
| 99.9999% (“six nines”) | 31.5 seconds | 2.59 seconds | 0.605 seconds |
* For monthly calculations, a 30-day month is used.
'Course it matters! It matters almost as much as if the downtime was expected or not.
Let's go for 9 fives!
Of course it's important w.r.t. availability, though four nines may not be enough for really critical applications. However, seeing that Stratus (not to be confused with Stratus Technologies btw) also promises four nines for performance and security makes me wonder… What is a 99,99% SLA for security supposed to be? Or performance?
Hoff- even back in my days at Interliant in 98 to 2000 (early ASP), 5 nines was the standard that our SLA's called for. 4 nines just wouldnt cut it. I think Microsoft was promising 4 nines if you ran NT or something like that.