|
i can handle if there is a dirrector failover so client connections will be
reset. In my setup it's not a problem, so i don't concieve it as one. I
believe that there are many other setups that this assumption applies to. In
these setups running stateful iptables is acceptable.
----- Original Message -----
From: "Joseph Mack" <mack.joseph@xxxxxxx>
To: <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Sent: Friday, September 27, 2002 4:40 PM
Subject: Re: RedHat 7.3 and lvs for kernel 2.4
> Alex Kramarov wrote:
> >
> > it's a wrong assumption.
> >
> > I run iptables and ipvs on a master-master cluster for months, without
any
> > problems.
> > I use stateful iptables rules, and i don't use ipvs sync daemon - these
two
> > negate each other.
>
> actually they are two independant problems. Currently iptables has no
> (stateful) failover capability. Harald Welte is working on it, but it
probably
> won't be done for a year. This means you can't have stateful iptables
rules
> on a machine that gets failed out and expect the connection to continue.
> If you want to failout a director, the iptables rules on the director must
> be stateless. If you want stateful iptables rules, you can have them on
> the realservers. If they failout, you loose the connection to the client
> in any case.
>
> Joe
>
> --
> Joseph Mack PhD, Senior Systems Engineer, Lockheed Martin
> contractor to the National Environmental Supercomputer Center,
> mailto:mack.joseph@xxxxxxx ph# 919-541-0007, RTP, NC, USA
>
> _______________________________________________
> LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
> Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
> or go to http://www.in-addr.de/mailman/listinfo/lvs-users
>
|
