[Linux-HA] about the resource stickiness

[Andrew Beekhof]

On 10/6/06, TORRESANI, Roberto <roberto.torresani at unitn.it> wrote:
> Well...I tried ptest and, really, after some -V it puts out some very interesting information. ;-)
> The only thing that remains obscure to me is how the colocation constraints are involved in the
> node's score calculation.
> Having:
> - heartbeat 2.0.7
> - a 1st resgroup (mysql) made of 2 res
> - a 2nd resgroup (apache) made of 4 res
> - a colocation constraint between them
> - resource_stickiness=50
> - resource_failure_stickiness=-100
> the total score shown by ptest is 1900 (I suppose it would be 1900/2=950 with the patch), so:
>
> mysql: 50 x 2 = 100
> apache: 50 x 4 = 200
> mysql + apache = 300
> ...and the rest of the score?
> How is it calculated?
>
> I'm sorry, I can't include the CIB in this moment (I'm at home), but I attach the interesting part of ptest output.


when you can post your CIB i'll have a look