I read this recently on a well known European SaaS accounting package :
“Our hosting center are upgrading their high-voltage power supply on May 18th between 00:15 and 06:00 (CET). In this period [the service] will be unavailable, this means that neither the web site nor the API will be available. So if you have any automatic jobs running within this time interval, they will not be run on May 18th.
As the system comes online again, we will be monitoring it to make sure that everything is running correctly.”
You what?! Down for nearly 6 hours – no website, no API, that’s absurd. I would have thought that these guys would have had a more resilient set up.
We use Rackspace – and have had less than 1 hour unavailability in 2 years. They rule.
