LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

[lvs-users] LVS FULLANT perf test result

Subject: [lvs-users] LVS FULLANT perf test result
From: emilio.campos.martin@xxxxxxxxx (Emilio Campos)
Date: Sun, 3 Apr 2016 20:28:53 +0200
F5 as another vendors make tuning in TCP stack, if you have some idea to
compare lvs vs f5 I think it will not be very conclusively.

In case you want to compare with F5, F5 *fastL4 *Profile would be the most
similar to LVS.

Regards!


2016-03-24 17:35 GMT+01:00 Jamie Dahl <jamied at meatball.net>:

> and to also note. the F5 use asics to accelerate a lot of L7 functionality
> and ddos offload; cots cannot really compete with that.   I am a fan of
> putting the heavy lifting intelligence on a proxy tier behind the LB since
> that allows seemless horizontal scaling.  very much a fan of L4/RR/dr or
> l3dr
>
>
>
> this was pounded out on my iPhone and i have fat fingers plus  two left
> thumbs :)
>
>
> > On Mar 24, 2016, at 01:07, Malcolm Turnbull <malcolm at loadbalancer.org>
> wrote:
> >
> > As no one else has replied I thought I would.
> >
> > LVS fullNAT is not widely used but is likely to perform very well as it
> is
> > simple layer 4 routing.
> > Most people use LVS for layer 4 dr/masq mode and HAProxy for anything
> that
> > needs Layer 7 i.e. a proxy (fullnat).
> >
> > F5 is the top of the range commercial product so neither LVS or HAProxy
> are
> > anything like it. However they can give 95% of the performance/features
> > when configured correctly.
> > Loadbalancer.org, Kemp & Barracuda Networks all use LVS for layer 4 load
> > balancing in their commercial products.
> >
> > You are much better off looking at what your requirements are first NOT
> > comparing to a random commercial product.
> >
> >
> >
> >
> >> On 23 March 2016 at 12:10, wei wang <lnykww at gmail.com> wrote:
> >>
> >> Hi.
> >> Is there any performance test result report for  lvs fullnat? I want
> >> to compare it with F5 hardware load balancer,so if there are some
> >> report, please send to me. thanks.
> >>
> >> _______________________________________________
> >> Please read the documentation before posting - it's available at:
> >> http://www.linuxvirtualserver.org/
> >>
> >> LinuxVirtualServer.org mailing list - lvs-users at LinuxVirtualServer.org
> >> Send requests to lvs-users-request at LinuxVirtualServer.org
> >> or go to http://lists.graemef.net/mailman/listinfo/lvs-users
> >
> >
> >
> > --
> > Regards,
> >
> > Malcolm Turnbull.
> >
> > Loadbalancer.org Ltd.
> > Phone: +44 (0)330 380 1064
> > http://www.loadbalancer.org/
> > _______________________________________________
> > Please read the documentation before posting - it's available at:
> > http://www.linuxvirtualserver.org/
> >
> > LinuxVirtualServer.org mailing list - lvs-users at LinuxVirtualServer.org
> > Send requests to lvs-users-request at LinuxVirtualServer.org
> > or go to http://lists.graemef.net/mailman/listinfo/lvs-users
>
>
> _______________________________________________
> Please read the documentation before posting - it's available at:
> http://www.linuxvirtualserver.org/
>
> LinuxVirtualServer.org mailing list - lvs-users at LinuxVirtualServer.org
> Send requests to lvs-users-request at LinuxVirtualServer.org
> or go to http://lists.graemef.net/mailman/listinfo/lvs-users
>



-- 
Load balancer distribution - Open Source Project
http://www.zenloadbalancer.com
Distribution list (subscribe): zenloadbalancer-support at lists.sourceforge.net

<Prev in Thread] Current Thread [Next in Thread>
  • [lvs-users] LVS FULLANT perf test result, Emilio Campos <=