[ClusterLabs] Antw: Re: Antw: Re: reproducible split brain

Ulrich Windl Ulrich.Windl at rz.uni-regensburg.de
Mon Mar 21 09:29:54 UTC 2016


>>> Dennis Jacobfeuerborn <dennisml at conversis.de> schrieb am 19.03.2016 um 15:10 in
Nachricht <56ED5DC4.9080100 at conversis.de>:

[...]
> I think the key issue here is that people think about corosync they
> believe there can only be two state for membership (true or false) when
> in reality there are three possible states: true, false and unknown.

Aren't there even more like "integrating", etc.?

> 
> The problem then is that corosync apparently has no built-in way to deal
> with the "unknown" situation and requires guidance from an external
> entity for that (in this case pacemakers fencing).

Isn't a node either member of a ring or not? Does "unkown" mean "temporariliy inconsistent", or what's the exact definition (now that we started talking on it)?

> 
> This means that corosync alone simply cannot give you reliable
> membership guarantees. I strictly requires external help to be able to
> provide that.
> 
> Regards,
>   Dennis
> 
> 
> _______________________________________________
> Users mailing list: Users at clusterlabs.org 
> http://clusterlabs.org/mailman/listinfo/users 
> 
> Project Home: http://www.clusterlabs.org 
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf 
> Bugs: http://bugs.clusterlabs.org 








More information about the Users mailing list