Here is our review of popular High Availability
technologies. It should be helpful in selection of the best solution for your
business.
Using high quality brand name servers in cooperation with redundant networks

On this diagram you can see the structure of data center that has
reservation for every component of the network. This reservation
includes multiple connection to the upstream ISP, multiple redundant routers,
multiple redundant switches, etc.
So, the possibility of network failure is minimized. The possibility of server
failure is also minimized by using high quality brand name servers.
Back side of this solution: very high costs, vulnerability from data-center problems.
Using Local Cluster Systems

On this diagram you can see the structure of local cluster system. It's principle of work is:
- in the normal mode all data from the customers servers are mirrored to the backup servers.
- in case of server failure all functions of failed server transfers to the backup server. From the customer's
point of view it looks like usual server reboot.
Using this solution you are independent from a server's failure.
The only vulnerability of this solution is dependence from data-center problems.
Using Global Cluster Systems

On this diagram you can see the structure of global cluster system.
The principles of work is similar with local cluster system but the backup servers
are located in another data-center. By using this solution you are independent from
both server and data-center problems.
From the our point of view this system does not have any lacks.
Our company will be happy to provide you with local or global cluster-based solutions.
|