Disaster Recovery Overview
Welcome to our Disaster Recovery overview article. In this article, we will talk about our disaster recovery procedures, systems and datacenter procedures.
What is a Disaster?
In the world of computers and web hosting, a disaster can be classified as anything that would interrupt service to our valued customers. This includes (but is not limited to) network outages, loss of data, power failure, and “acts of god.”
A network outage is when part of the network malfunctions and causes either intermittent connectivity or complete lack of communication between the servers and the outside world. This normally happens when a switch, router, network cable, network card or other piece of networking equipment fails.
We work with datacenters that have very high standard Service Level Agreements (SLA) that include timely network malfunction repair and equipment replacement policies.
Hard Drive Failure/Data Loss
Data loss happens when multiple hard drives malfunction or a RAID controller fails and data is no longer accessible, the data becomes corrupted and/or a RAID is at risk of failing, or has failed (this is very rare).
Again, we only work with datacenters that have a high standard SLA regarding faulty equipment replacement so that we can remove/replace bad hard drives, RAID controller and other hardware (motherboards, CPU, memory, etc.) that may cause data lose downtime in a timely manner.
In most situations, if a hard drive fails, a server will remain online due to our RAID configuration and redundancy setup, thus allowing us time to contact the datacenter to have the drive replaced all while the server is still online.
In the rare situation when multiple drives malfunction or the RAID controller fails, we contact the datacenter to have the faulty hardware replaced so that we can begin the data restore process. We have very sophisticated backup procedures and systems in place to help minimize the possibility of data loss and to help expedite data restoration for our customers.
All of our services (excluding VPS clients who have opted out of our backup solution) include daily, backups that are stored on dedicated backup machines at the datacenter. For our VPS clients, these are incremental backups, for all of our other services, we make full account backups. We keep these backups for up to ~7 days before removing the oldest backup to make room for new backups. This means at any given time, we have 7 days’ worth of backups for all of our clients/services.
In a full RAID failure situation where the entire server needs to be restored (this is VERY rare), we take the newest backup and begin the restore process. This process does take time as there is a large amount of data that needs to be moved from one server to another. Because of this, the exact time needed to restore a server can change from server to server and location to location.
We only work with datacenters that have power backups in place. Most locations have an uninterruptible power supply (UPS) and a diesel generator(s) to prevent power loss in a long term power outage situation. For more information regarding power backup systems, please scroll down and visit our location specific pages.
Acts of God / Natural Disasters
Natural disasters are always a possibility, depending on the location. For example, our Australian Datacenter was created from the ground up to withstand flooding and has tools to prevent any damage to the building, servers, power grid and more. In comparison, our San Jose datacenter is built with earthquakes in mind, and has the infrastructure to minimize the impact of any earthquake
Below you will find links to articles that cover each location in more detail.