On Wed, 2007-12-19 at 11:52 +0100, Alain Moulle wrote: > Hi Lon > > Many thanks for your last both responses, it's much clearer > in my head now ... ;-) > > When you speak of "a method to enable master-set-wins (e.g. for > operation w/o heuristics)" , will it be delivered by a patch when > ready ? or just a matter in configuration file ? It will be both. The current idea is to use a quasi-"fence agent" which talks to qdiskd. Basically, "fence_qdisk" asks qdiskd "Am I in the master set?" - if we are not, qdiskd "fencing" fails immediately. This "fencing" operation needs to happen before other fence agents. If we /are/ in the master set, we flip the "dead" node/partition's state on the quorum disk S_EVICT (Evicted) - which will (probably) cause the node to die, and return success to fenced. At this point, normal fencing occurs. This provides also a quick "not supported for production use without other 'real' fencing" test agent, too... The limitation is that you cannot use multiple fence methods - or if you do, you must reference the fence_qdisk agent from all methods. > if patch, when do you think it could be available ? because > I'm interested ... I guess it will be sort of a way to establish > that one of the two-nodes cluster is master and in case of > "perfect symmetrical splits", it is always the master the winner ... > if so, that will be fine for me . Yes, that's the intent - I'm also thinking about methods of how to simplify things somewhat. If you have input, let me know. :) -- Lon -- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
