[Linux-HA] HP DL380 x2 and one raid

[Peter Kruse]

Lars Marowsky-Bree wrote:
> On 2005-08-17T09:09:18, Peter Kruse <pk at q-leap.com> wrote:
> 
> 
>>
>>1. Some failure to make the cluster want to reset a node
>>2. no heartbeat from the node
>>3. no heartbeat from the stonith device
> 
> 
> ... all of these are quite commonly caused by a single error, namely one
> node losing (all) network connectivity.
> 

That's not the point, the cluster already has decided to reset one node.
I'm not talking about how to get to this decision.  I'm talking about
when can stonithd return success if both the node and the stonith device
are unavailable?

	Peter