[Planetlab-devel] [Fwd: disable slice creation mechanism]

[Thierry Parmentelat]

IMHO, not only is it not too hard, but in addition it sounds like the 
enabled flag for users/persons;
and like for persons, during the site creation process, there is a need 
to have sites be put 'on hold' until they actually have their nodes up 
and running. It would be more explicit to have the site disabled than to 
have its max_slices to 0.
So globally I think 2) is going to be very familiar as compared with the 
rest of the system.
-- Thierry

Stephen Soltesz wrote:
> This probably belongs on the devel mailing list.
>
> ------------------------------------------------------------------------
>
> Subject:
> [pupadm] disable slice creation mechanism
> From:
> Stephen Soltesz <soltesz at CS.Princeton.EDU>
> Date:
> Wed, 13 Jun 2007 15:42:39 -0400
> To:
> PU PlanetLab Admin List <pupadm at lists.cs.princeton.edu>
>
> To:
> PU PlanetLab Admin List <pupadm at lists.cs.princeton.edu>
>
>
> Looking at the existing monitor code, and the (evidently new) policy 
> of assigning sites a different number of max_slices, there is a 
> mis-match.
>
> Monitor currently disables slice creation by setting the max_slices to 
> zero.  However, upon re-enabling a site, there is no way to 
> differentiate a site (that I see) as being in the max_slices==2 group 
> (or any other group we may define in the future) or the max_slices==10 
> group.
>
> This suggests at least two options: 1) have a site attribute that 
> distinguishes the member class of a site (and hence it's max_slices 
> value), or 2) provide an 'enable/disable' flag for slice creation at a 
> site.
>
> Faiyaz and Tony felt that 2) would not be too hard.  Other thoughts?
>
> Stephen.
> ------------------------------------------------------------------------
>
> _______________________________________________
> Devel mailing list
> Devel at lists.planet-lab.org
> https://lists.planet-lab.org/mailman/listinfo/devel
>