Sessions are useful without session failover. Many applications won't work at all without sessions. Failover will be handled by starting a new session on another node; something which may be perfectly acceptable.
As for the performance problem: building a bigger bottleneck only gets you so far; it doesn't solve the fundamental problem. Again, that may or may not be acceptable.
Re:How does this work with sessions?
Posted by: ulric on June 14, 2005 03:18 PMAs for the performance problem: building a bigger bottleneck only gets you so far; it doesn't solve the fundamental problem. Again, that may or may not be acceptable.
#