AFC will use commercially reasonable efforts to make our services available with an annual uptime of 99.9% during the service year. Click here for a copy of our SLA document.

In the event that we are not able to meet this commitment you will be eligible to receive a credit as described below.

TERMS

• A credit will be applied to the next months services.
• Credit will only be issued if the calculated credit amount is greater than $100.
• The service year is the preceding 365 days from the date of an SLA claim.

EXCLUSIONS

• The credit does not apply if the outage is due to user supplied equipment.
• Caused by factors outside of our reasonable control, including internet access or related problems beyond the demarcation point of AFC.

UPTIME - 100% - Is it really possible?

The key to uptime is multiple servers in multiple locations. How many nines can you expect? It important to examine exactly what those percentages work out in terms of actual minutes and seconds.

Nines	Uptime %	Downtime %	% Downtime Sec	Downtime/yr
1	90.0000%	10.0000%	250000	3 days
2	99.0000%	1.0000%	25000	7 hours
3	99.9000%	.1000%	2500	42 minutes
4	99.9900%	.0100%	250	4 minutes
5	99.9990%	.0010%	25	30 seconds
6	99.9999%	.0001%	2.5	2.5 seconds

- 99% uptime (7 hours) is enough time to replace a failing drive and restore from backup tape.

- 99.9% uptime (42 minutes) is tough unless you have 24x7 staff to resopnd to needs. This is only enough time to troubleshoot a hardware issue and replace/failover if you have the parts on hand.

- 99.99% uptime (4 minutes) is the most drastic jump. In 4 minutes there isn't much you can do in terms of troubleshooting and repairs. This isn't even enough time to reboot a server. At this point you have to have unmanned management of your server and network issues and have failover hardware. Your monitoring has to automatically reboot services or servers. This is where load balancing, replication, and clustering are essential and be in at least 2 datacentres.