High availability by virtualization? Try RDBMS

I am considering VMware for high availability, but I am not sure about all that VMware offers. I have a Progress database on a UNIX server with hundreds of windows clients and web applications attaching to it through Progress Brokers via service file ports. I need to provide 365/24/7 uptime and across at least two servers, sort of like an Oracle Rack. This would allow me to upgrade the OS, reboot a server or take a server down for maintenance without affecting the database or the users. With our new web business, East and West coast facilities and vendors managing our stock and replenishment, we need to be available all of the time. So far I have only found solutions that will give me a 2-5 minute down time between switching from one server to another. Do you have any recommendations?

Virtualization software, such as VMware, will not provide the functionality want, as it is not a high availability system. In addition, most high availability solutions will usually have downtime associated with it, which is why it is called high availability and not fault tolerance. Progress has a number of systems that provide some level of 24/7 availability through its Sonic messaging architecture, or what they refer to as Sonic Continuous Availability Architecture (CAA). Try looking at this first to see if it might meet your needs.

Also, take a look at the high availability that is offered within the relational database management system (RDBMS) itself. I have yet to use this offering, but as one of the pioneer users of Progress back in 1987, I can vouch for the strength of its offerings. High availability can be provided for with failover cluster support and site replication through hot backup and replication. Some Progress features support this: automatic crash recovery, transaction logging, roll-forward recovery, point-in-time recovery and site replication and two-phase commit protocol. The two-phase commit protocol maintains the integrity of data that span multiple sites and allows a standby machine to be maintained at either the same site or at a different location for business continuity purposes. Look at this RMDBS datasheet from Progress that provides for more information.

For other Linux-based high availability and/or clustering solutions, check out Beowulf and the Linux HA project.