Basic and advanced src-ip balancing

Question

What is the 'key' difference between the basic src-ip method and the advanced sticky src-ip methods?
I can understand in the advanced one, CSS actually "sticks" a client to a server based on the client's source IP address.
I can understand the basic one as well, CSS directs all client requests coming from the same source IP address to the same service. So, this method too provides a bit of stickness.
The question is how does they differ from each other, in what detail.
Got any clue.
thanks

Answer 1

the sticky method (advanced) creates a sticky table, each entry containing client-ip/server-ip. When a server goes down, the entry is removed and a new one created.
The standard method does not use a sticky table but a hash algorithm. The algorithm takes the client ip and returns the index of a server.
The same client ip returns always the same index.
If a server goes down, the algorithm select the next available index.
This is where is the difference.
Assume the hash algorithm returns index of server-A. While A is alive, everything is fine. When A goes down, traffic will be sent to the next index, let's say B.
When A comes back alive, traffic is again sent to A.
With the advanced method, since a new entry pointing to B is created when A is down, when A comes back up, traffic is still sent to B.
Another difference is if you add or remove servers from the list. The hash algorithm may return different index.
Gilles.