[Openais] Split brain when using EVS library

Steven Dake sdake at redhat.com
Sun Sep 14 13:01:04 PDT 2008


On Sat, 2008-09-13 at 21:05 +0200, Lars Marowsky-Bree wrote:
> On 2008-09-13T20:58:26, Ruppert Koch <ruppert at rcsc.de> wrote:
> 
> > Fault detection as well as membership are managed by the Totem protocol. 
> 
> Right.
> 
> > I assume the following happens:
> > 
> > A node P experiences a token timeout.
> 
> Ah, and this is the point which I'm curious about - why does this occur?
> I assume we're running mlocked sched_rr etc too, and the timings
> shouldn't be that sensitive - so where does the timeout happen, and why?
> What gets stalled?
> 

pretty sure there was a bug relating to this matter in 0.83 which is
fixed in the latest tip of whitetank.

Regards
-steve

> 
> Regards,
>     Lars
> 



More information about the Openais mailing list