difference between persistence on VS and round robin on the pool.

Round robin is the load balancing method i.e. connection 1 to pool member 1, connection 2 to pool member 2 etc...and then back again to pool member 1/pool member 2. If you don't have persistence and a returning client accesses the VIP then the load balancing method will kick in and it'll be sent to the relevant pool member accordingly. If you have persitence set (as you mention above) then a returning connection (within the timeframe setup) will bypass the load balancing method and the connection will be put through to whichever pool member you were persisted to originally.

Hope this helps,

N

Think of them as actions at two different times.

Round-Robin load balancing method describes how the poolmember will be chosen for a connecting coming through to the pool where no existing session exists. i.e. The poolmembers will be chosen on a strict RR basis (No account being made for number of concurrent connections, response time etc).

The persistence describes an over-ride for that load balancing method (round robin) that tracks the source IP address and treats all connections from a given IP address as a single 'session' and ensures that subsequent connections (Subject to a timeout) will be sent to the same poolmember.

RR does not guarantee that your poolmembers will receive the same load or the same number of connections. They will however receive the same number of sessions...

H

The persistence on the VIP will be checked first. So source persistence but different source ip addresses will mean it doesn't match so the load balancing algorithm will be used.