|
Yunfeng Hou wrote:
> I checked the mailing list and did not find something
> such as roadmap or function list of next release.
The question of a roadmap came up in the early days of LVS
and Wensong said there would not be one. At the time
Wensong told me that he regards LVS as his personal
project. It is not a large internet project, like the
kernel, where anyone can submit code. I know I'm speaking
for someone else here and Wensong might come up on the
mailing list and say more.
Still there is no clear statement on the policies involved here.
It would be nice to have them all public so people who
want to write code for LVS know what they have to do
to get their code accepted.
The best thing to do if you want to write code for LVS is
to make it a loadable module that will operate independantly
of the main code. At least this is the procedure people are
using so far.
There is lots of useful patches out there for LVS that are not
accepted in Wensong's code. Unfortunately they are not easy
to maintain with the newer versions of LVS and so don't get
used as much as they could.
> 1. user defined scheduler: user can write their own
> scheduler to guide the load dispatching
you can already do that. Take a scheduler module and
rewrite it to schedule your way.
> 2. session persistent extend: I noticed that 3 methods
> to extend a session in the mail
> http://marc.theaimsgroup.com/?l=linux-virtual-server&m=104467069124909&w=2
> And I prefer the first one to extend it again with
> the persistent time. So maybe all these implemented
> and have control variable to choose one.
I'm sure you could do that if you wanted.
> 3. DOS attack: add functions such as BIG-IP syn check.
> http://www.f5.com/solutions/tech/security/
again there's no reason why you couldn't write that either.
Horms maintains ipvsadm. If you write any of the above code,
you can send patches to control your code to Horms. Wensong
has the final word on what goes into ipvsadm, so you should
check with Wensong first to see if he will allow your patches.
There is not much point in writing code that will not be
maintained with the main LVS code stream, unless you have
a large interest in maintaining it yourself.
Joe
--
Joseph Mack PhD, High Performance Computing & Scientific Visualization
LMIT, Supporting the EPA Research Triangle Park, NC 919-541-0007
Federal Contact - John B. Smith 919-541-1087 - smith.johnb@xxxxxxx
|
