Thats right we are trying for less then 0.8 seconds of down time per day.

Lets face it customers don’t care if its your fault that they cant reach you… they just cant reach you… They dont know if its your server or a router or your firewall or your internet WAN provider… all that they see is a service that cannot be connected to…

So…

Lets make some assumptions, average ping 50ms, from anywhere to anywhere, thats a 16 packet window… if only one user was utilizing the “stream”… thats insane… how do you guaranty that you wont lose 16 packets. Now I know any Company that thinks it needs a 5 9 service is going to have more then ONE user at a time…

So maybe 16 packets is so small that customers wont notice… but it still happened… and you’ve already blown your 9′s…

To often small and medium companies focus on making sure that their back end and servers have the 9′s but forget to factor in the rest of the world, the tubes that connect us all are not under our control… accidents and congestion happens… (occasionally black holes appear the suck up all the traffic from Youtube). Unless you’ve got the power and cash of Google your never gonna reach more then 3 9′s… if you’re moving a Million USD a minute then maybe you have a business case…

Downtime is never acceptable… but ask your self can you afford to be truly dressed to the 9′s?