Andrew Beekhof wrote:
On Apr 24, 2006, at 2:35 PM, Lars Marowsky-Bree wrote:
On 2006-04-24T14:21:28, Andrew Beekhof <[EMAIL PROTECTED]> wrote:
and stopped on everywhere else.
not as easy (as in we'd need to add something to the PE)
if these nodes would use '0' or no value at all (we map that to
delete) then we could do it now.
Well, hard to do. As they have no idea to know whether they are in the
largest group or not. If they knew the answer to that question, we would
have already solved the problem ;-)
nod, i know its wishful thinking on my part :-)
but some sort of #is_max operation for expressions wouldn't seem too
difficult.
we just need to sort out the N+N split and we're golden :)
Computing the best-connected set of nodes in the subset is an n-p
complete problem :-(
We have a heuristic for it in the CCM.
It's not the only answer, but it's very effective.
Having alternative answers with different characteristics seems good...
The membership idea is admittedly pretty drastic.
--
Alan Robertson <[EMAIL PROTECTED]>
"Openness is the foundation and preservative of friendship... Let me
claim from you at all times your undisguised opinions." - William
Wilberforce
_______________________________________________________
Linux-HA-Dev: [email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
