Re: [cf-env] [abacus] Changing how resources are organized

Piotr Przybylski <piotrp@...>

The main aggregation needed is at the resource type, however the aggregation within consumer by the resource Id is also something we would like to access - for example to determine that application used two different version of node. Instead of introducing resource type, the alternative approach could be to augment the consumer id with the resource id. 


-----Jean-Sebastien Delfino <jsdelfino@...> wrote: -----
To: "Discussions about Cloud Foundry projects and the system overall." <cf-dev@...>
From: Jean-Sebastien Delfino <jsdelfino@...>
Date: 12/09/2015 11:51AM
Subject: [cf-dev] Re: Re: Re: [cf-env] [abacus] Changing how resources are organized

It depends if you still want usage aggregation at both the resource_id and resource_type_id levels (more changes as that'll add another aggregation level to the reports) or if you only need aggregation at the resource_type_id level (and are effectively treating that resource_type_id as a 'more convenient' resource_id).

What aggregation levels do you need, both, or just aggregation at that resource_type_id level?

- Jean-Sebastien

On Mon, Dec 7, 2015 at 3:19 PM, dmangin <dmangin@...> wrote:
Yes, this is related to github issue 38.

We want to group the buildpacks by a resource_type_id and bind resource
definitions to the resource_type_id rather than to the resource_id. However,
when we make this change, how will this affect how abacus is doing all of
the calculations. The only change that I can think of is for abacus to use
the resource_type_id rather than the resource_id when creating the reports.

View this message in context:
Sent from the CF Dev mailing list archive at

Join to automatically receive all group messages.