You rely heavily on email to operate your business and as a result AttainResponse has gone to great extent to ensure that your email is available 24x7, 365 days a year. To maintain an industry-leading 100% uptime guarantee, we have chosen a strategy popularized by Google.

 

Essentially, a large numbers of computer servers built on commodity hardware (see article: Google Revealed). As you know, server failures are common, so as you will see Google, and AttainResponse, arranges servers in pairs or clusters, then builds failure detection into the network or the software.

 

Service Architecture

Each of our four core email services operates on their own highly redundant, load-balanced cluster of servers. The four services include:

• Inbound email (SMTP)
• Outbound email (SMTP)
• Remote email client access (POP & IMAP)
• Web-based email access (webmail)

If any server in a cluster fails there is always additional servers standing by to instantly take over. This way there is no noticeable degradation in service performance.

 

Email Storage
Email storage is spread across a large number of mirrored pairs of servers. If a primary server becomes unavailable the backup server immediately assumes the responsibility until the primary is restored.

 

Databases
Database scaling is implemented using MySQL master-slave replication. All slave MySQL servers have a live copy of all data, and therefore any server can take over the role as master in the event that the master server fails.

 

DNS
DNS is also designed for high availability through a load-balanced cluster of caching name servers used to answer all DNS queries.

 

Spare Servers
Each server rack has at least one spare server, which is fully configured with OS and application software. The spare is standing by and ready to take over for any server on the rack at any time. Switching over a spare server's identity is nearly 100% automated and can be accomplished within minutes.

 

Data Center
Our equipment is hosted with top tier provider The Planet. Highly available, diverse bandwidth is available through several tier-one network carriers including WilTel, SBC, Sprint, AT&T, Qwest, Time Warner Telecom and MCI.

 

The Planet data centers offer a 100% network uptime guarantee (SLA) and have industry leading Power, Cooling, Security, Fire Suppression and Technical Support.

 

Scalability
The infrastructure of AttainResponse was designed and developed to quickly and smoothly scale out system capacity to meet demand. When new servers are needed, pre-built software configurations can be loaded in about 20 minutes and the new servers can be seamlessly added to clusters wherever they are needed.

 

Adding email storage capacity is as simple as bringing online new pairs of email storage servers.

 

All of the core email services and back-end databases can all scale independently from one another. For instance, inbound email (SMTP) servers are required to do intensive spam identification and processing as well as two layers of anti-virus filtering. This efficiency in using server resources allows us to stay competitive on price while offering superior service quality.