On Sun, Apr 08, 2007 at 11:54:42PM +1100, Russell Coker wrote: > On Sunday 08 April 2007 19:14, Russell Coker <[EMAIL PROTECTED]> wrote: > > The ha.cf file already lists all nodes that are in the cluster via the > > "node" directive. Surely if a node calling itself "foo" asks to join the > > cluster then regardless of whether it has a suitable auth key it should not > > be accepted if the list of valid nodes includes no "foo". > > Apr 8 22:52:35 ha2 heartbeat: [2929]: WARN: string2msg_ll: node > [ha2-unstable] failed authentication > Apr 8 22:52:35 ha2 heartbeat: [2929]: WARN: string2msg_ll: node > [ha1-unstable] failed authentication > Apr 8 22:52:35 ha2 heartbeat: [2929]: WARN: string2msg_ll: node > [ha1-unstable] failed authentication > Apr 8 22:52:36 ha2 heartbeat: [2929]: WARN: string2msg_ll: node > [ha2-unstable] failed authentication > > For even more annoyance I get the above messages repeatedly in my syslog when > I use different auth values.
I believe that the work-around of choice is to run different clusers on different ports. > Running two clusters on the same VLAN is not going to be viable until after > this bug is fixed. Do you plan to run a back-ports repository for newer > versions of Heartbeat on Etch after the release of Etch? If you are > considering such things then this makes a good reason IMHO. Yes, I usually make newer versions available in backports as they are released upstream. -- Horms H: http://www.vergenet.net/~horms/ W: http://www.valinux.co.jp/en/
