[Linux-cluster] Dependency graphs, order of startup and failover handling - kill the 2nd node in 2-node cluster without quorum

[Klaas]

Hi Eftychios;

On Fri, 2006-12-08 at 13:08 -0500, Lon Hohberger wrote:
> So, the short answer to your question is, "Well, not yet, but, er, funny
> that you should ask ..."
> 
> On Thu, 2006-12-07 at 22:11 +0200, Eftychios Eftychiou wrote:
> > Got a few questions regarding RHCS. Would be grateful if someone could
> > answer them.
> > 
> > 1. Fencing in a two node cluster according to the FAQ could lead to
> > each node trying to fence each other resulting in total cluster
> > blowup. The suggested solution of that problem was to use serialized
> > fencing devices. 

Being new to *this* type of clustering,
I had to delve into this matter too (see my other thread).
As far as I understand,
this so called STONITH behaviour is perfectly OK!
As long as 1 of the two nodes is slightly faster [fencing
the other] nothing is wrong. You end up with a 1-node
cluster which has quorum and will run your services.

Everything is better than accidentally having two nodes
running the services.

Regards,
Klaas