Question:
I am using Oracle fail safe and
I need to ensure that there is no downtime if one of the nodes dies.
How to I achieve 100% continuous high availability in Fail Safe?
Answer:
There is no such thing as high
availability on a Windows Fail Safe platform, and Most CIO’s will
laugh out loud at the suggestion of Oracle Fail Safe as a viable high
availability solution, not because of Oracle, but because of the long
history of Windows disasters.
Ask yourself, would
you deploy a mission critical database on a platform that has a
history of vulnerabilities, hacking, and requires weekly patches from
the vendor? Oracle Fail Safe was originally designed for
networked personal computers and it does not scale up well.
Fail Safe is also
problematic because there is a huge outage when the failover instances
are started, several minutes of unnecessary downtime, as Fail Safe
starts the instances and listener on the failover server.
This delay in Fail
Safe makes it unacceptable in any true high availability environment.
In a real-world environment, you are better off to quickly re-mount
the disk to a UNIX server, and re-start the server, using the concept
of the
Database Area Network.
The Database Area
Network is more flexible than Fail Safe
