Ip_hash clarification

Hi,

i need details regarding ip_hash…

if i have client IP1: 192.168.1.1 and client IP2: 192.168.1.2

and I have 2 servers on the upstream… s1 and s1… with IP_Hash
configured

if client IP1 accesses the site it will go to lets say S1 would it
actually
mean that the next request coming from a new client IP Client2… will it
go
to s2?

finally how do they compute where to throw the request…

tia,

On Thu, Feb 25, 2010 at 11:40 AM, J M [email protected] wrote:

mean that the next request coming from a new client IP Client2… will it go
to s2?

I believe ip_hash uses only the “Class C” part of the address. This is
so people with firewalls that have them “hop” IP addressess still get
directed to the same back-end. So since both clients are coming from
192.168.1.XXX, nginx will direct them to the same back-end server.

I was actually considering a patch to make this configurable a few
moths ago, but we decided to go with session replication on our
back-ends instead, so I stopped looking into it.

For testing, you could put a client on 192.168.2.XXX to see the
desired behavior.


RPM

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.

| Privacy Policy | Terms of Service | Remote Ruby Jobs