We don’t normally discuss such things in public forums but in the two cases we’ll make an exception.
The OP is misleading in this post. We had the database backup from the night before and also had 4 weeks of incremental site file changes. We didn’t have the full set of files of their sites. It happened on a Saturday evening and we called everyone immediately and worked with the handful of affected clients to restore their sites from their local copies. We then applied our incremental changes and restored their databases.
What happened was a double drive raid failure on a particular node during the taking of the full backup. The backup software was removing the full and making a new when it happened. That’s a very rare event which we handled well, learned from and improved our infrastructure making it even more robust.
As for the downtime that is correct 5 hours in 5 weeks which totals 7 in 3 years. The first 2 hours over a year ago was a firmware issue in the TOR switches. The second and third events in that 5 week window were a large scale DDOS and an internal network attack by hacked customer accounts with weak passwords.
We’ve since switched over to new TOR switches, worked with our up stream providers to enable us to push our own null routes to the borders without having to call them and locked down customers SSH access by IP. Going back through our cancellation requests I do not see any cancellations listed due to downtime. I believe that’s a testament to our wonderful support and infrastructure that we offer for Magento store owners.
Hope that helps everyone make informed decisions about your hosting. If you have any further questions please contact us at