Amazon Web Services Get Redundant
They’re not in direct competition — yet — but Google is no longer the only A-list cloud services provider with disaster recovery and failover. To wit, Amazon Web Services has launched Multi-Availability Zone (Multi-AZ) deployments for Amazon Relational Database Service (RDS), giving users redundant cloud backups. Here’s the scoop.
Here’s the basic idea: if you opt in for a Multi-AZ deployment, your database gets replicated to a discrete and geographically distant data center with an entirely different infrastructure. This “standby” version of your database is provisioned automatically, and whenever you make a change it’s synchronized across availability zones.
In the event of downtime, whether planned (routine maintenance) or unplanned (earthquake), the Amazon RDS database fails over to the standby replica with no need for administrative intervention. Amazon Web Services says that the boost to availability and fault tolerance in Multi-AZ deployments means they’re perfect for critical production environments.
Disaster recovery and failover is something the cloud does really well, and it’s good to see Amazon Web Services figure out a way to leverage that in their database offering. My only concern: like Google, Amazon is probably going to have to deal with data loss due to user error being synchronized across the cloud, and there’s no word on how long it might take for them to catch up.