Hello everyone,
Please forgive my butting in, but I have a couple of questions about
using LVS, before I make a decision on which load balancing software to
attempt.
I have read the 'mini how to' by Joseph Mack, and appreciate the depth
of information covered in the document. I have not read the 'full-sized
how to' yet.
My current needs are to set up a load balancing solution for my QA team
that is both effective and inexpensive. Could someone point me towards
any kind of benchmarking data regarding the number of 'messages' or
packets were sent to x number of 'real servers' on a particular hardware
configuration?
The machine I intend (only system currently available) to use is a P3
450 w/256MB ram. I need to configure this machine to load balance for
two sets of real servers. Is this plausible?
I have 5 apache servers with the above listed hardware configuration. I
also have 2 Sun Enterprise 450 servers fro the 'application' servers. I
will need to be able to adjust the number of apache servers up or down
'on the fly' for the QA team. Will the balancer have to come down for
this change or does it have a means for altering the configuration in
real time?
Will the LVS system support the Sun application servers? I noticed in
the 'mini-how-to' that there is a script to be executed on each of the
'real servers'. Will this script function correctly in Solaris 8?
Thank you all for your time and your answers. I hope to be suing your
product soon.
Daniel Curry
Systems Administrator
CGtime, Inc.
625 Second Street
Suite 201
San Francisco, CA 94107
ph: 415-348-6516
fx: 415-348-6505
cell: 510-579-6680
|