LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

A new 2.2 VS patch

To: linux-virtualserver@xxxxxxxxxxxx
Subject: A new 2.2 VS patch
From: Peter Kese <peter.kese@xxxxxx>
Date: Tue, 18 May 1999 22:05:13 +0200
Dear folks...

The Virtual Server patch development for the 2.2 kernel is
continuing. Not only, that there had been some strong discussions
about what to put into the comments of the files, some new code
had also been written.

I have decided to call it VS patch 0.2 for kernel 2.2.9 and
I sincerely hope not too many bugs have escaped. However I am
expecting some, as I always do when I keep changing the code
until the very last moment before the release.

Putting bugs to the side, my opinion is, that the 0.2 release
is mostly feature complete. The most important things added to
the code were the following:

- tunneling,
- TCP FIN packet detection,
- improved performance of the WLC scheduling algorithm, by
  ignoring expired (FIN) connections
- PCC scheduling algorithm

(When compiling the kernel take care of recompiling
 and reinstalling the ipvsadm as well.)


I would like to ask Wensong to fix the licensing,
acknowledgments and change log in the code and put the patch 
to the Virtual Server web site so everyone can access it. Feel
free to fix bugs too.

However I would like to stay the maintainer of the 2.2 patch
for some more time, because I do have some more ideas about
changing some things that weren't implemented in the best way.
Later on, I wouldn't mind sharing the maintenance with someone
else. If you have some more suggestions, please feel free to
write.

Also in the future I will try to do my best with responding
to my mail and keep the track with the Virtual Server mailing
list. I admit, I do have some problems by handling all the
activities (job, kernel development, studies...) and sometimes
can't bother by answering all the mail. I think the quarrel
about change log of the new patch was partly due to that reason.

And I still need to write the readme file!

Enjoy,
                        Peter


P.S.
PCC (persistent client connection) scheduling algorithm
needs some more explanation. When PCC scheduling is used, the
connections are scheduled on a per client base instead of per
connection. That means, the scheduling is performed only the
first time a certain client connects to the virtual IP. Once
the real server is chosen, all further connections from the
same client will be forwarded to the same real server.

PCC scheduling algorithm can either be attached to a certain
port or to the server as whole. By setting the service port
to 0 (example: ipvscfg -A -t 192.168.1.10:0 -s pcc) the
scheduler will accept all incoming connections and will
schedule them to the same real server no matter what the
port number is.

As Wensong had noted before, the PCC scheduling algorithm
might produce some imbalance of load on real servers. This
happens because the number of connections established by
clients might vary a lot. (There are some large companies
for example, that use only one IP address for accessing the
internet. Or think about what happens when a search engine
comes to scan the web site in order to index the pages.)
On the other hand, the PCC scheduler resolves some problems
with certain protocols (e.g. FTP) so I think it is good to
have it.

Attachment: ipvs-0.2-2.2.9.tar.gz
Description: GNU Zip compressed data

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