AW: Failover if internal NIC of active goes down ?
Fri, 9 Feb 2001 09:15:15 +0100
On Fri, Feb 02, 2001 at 07:28:48PM -0700, Alan Robertson wrote:
> Ralf.Staiger@pdb.sbs.de wrote:
> > Heartbeat monitors the link for gettin` the status, doesnīt it ?
> > So it would be a nice feature to:
> > 1) maybe monitor more than the link needed for the heartbeat
> > 2) start the failover if a link status is: dead
> > Maybe thatīs not the focus of the software, but the functionality is nearly
> > build in....!?
> > Again: I donīt want to monitor a service ! (just the link status of the
> > NICs)
> I agree. The steps in doing this are as follows:
> 1) Divide the heartbeat cluster/resource manager into a separate process.
> 2) Make the change involved.
> I expect to begin (1) within the next month.
Well if you are really differing between local/global resources the
resource model will get a bit more komplex and the need for a "new"
resource manager is not avoidable. Actually for DRF I am needing
such a piece too, cause I need to know what is going to be monitored.
So I have started writting a more complex one. Perhaps there is some
interest for it when I finish it (perhaps this month).