Home > Articles > Software Development & Management

  • Print
  • + Share This
Like this article? We recommend


For example, the availability targets of the system may be 99% of the time between the hours of 8 a.m. and 9 p.m., and 98% of the time between the hours of 9 p.m. and 8 a.m. In this scenario, it's important to ensure that backup routines are devised so that they can be carried out without making the system unavailable.

Such availability targets might affect the choice of operating system—lfor example, Windows NT with IIS caused problems for many system administrators who were under pressure to take down the system for maintenance as infrequently as possible. Microsoft has addressed this issue and claims 99.999% server uptime for the Microsoft Windows 2000 server family. The number of maintenance tasks that require rebooting of the computer—lsuch as installing hardware and software—lhave been drastically reduced (about time, most Linux fans would say).

Hardware reliability and redundancy would also greatly affect availability of an e-business system. The concern is not simply for server hardware, but also any firewalls, proxies, and routers that might stand between the end user of the system and the server. Full redundancy (failover) of all hardware is essential to maintaining high availability. Many larger companies effect this by using two completely separate sites—lone for the "live" site and one for the failover site, preferably using different ISPs to connect to the telecoms network also.

Within any one site, clustering (server and/or database redundancy) and RAID (disk redundancy) are methods used to improve reliability through redundancy of hardware.

  • + Share This
  • 🔖 Save To Your Account