--- On Thu, 10/16/08, Graeme Fowler <graeme@xxxxxxxxxxx> wrote:
> From: Graeme Fowler <graeme@xxxxxxxxxxx>
> Subject: Re: [lvs-users] failover on SH
> To: jeff.pang@xxxxxxxxx, "LinuxVirtualServer.org users mailing list."
> Date: Thursday, October 16, 2008, 4:39 AM
> On Thu, 2008-10-16 at 00:25 -0700, Jeff Pang wrote:
> > I have chosen SH as the scheduler algorithm on my LVS.
> > All run file, requests from client A will always go to
> realserver A.
> > But, when realserver A got died, client A's
> requests continues to go to realserver A.
> > I have two realservers, why LVS didn't redirect
> client A's requests to realserver B?
> > failover failed on SH algorithm?
> Are you using persistence?
Since we're using source-hash, why we need persistence?