[lvs-users] the SH scheduler configuration

To: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Subject: [lvs-users] the SH scheduler configuration
From: Alexis Berger <aberger@xxxxxxxxxxxxx>
Date: Mon, 23 Jun 2008 14:41:25 +0200
Hi all,

I have a big issue using LVS and I have been looking for information on 
the Internet for one day without any success. So I hope someone here 
could help me :)

Let me explain my situation and why I thought about using the SH scheduler.

I have a LVS pointing on 2 real servers.
This LVS is accessed from Internet by anyone, which may require for some 
requests persistence handling (login on a web page which grants access 
to other webpages).
So I used the WLC scheduler, with persistent set to 600.

Now I have a new problem to deal with: one external application performs 
a huge amount of requests (without persistence), using 2 differents IP 
If I do not change the scheduler, only one of my real server is used.

So I looked inside the documentation for a strategy allowing to define a 
specific IP address to  specific real server, and I read:
         sh* - Source Hashing: assigns jobs to servers through looking 
up a statically assigned hash table by their source IP addresses.

By the way, how to define the hash table and thoses mappings? In which 
files, which syntax, which parameters, etc.

Thanks in advance as I have no clue for the moment.
Alexis Berger

<Prev in Thread] Current Thread [Next in Thread>