thanks to ibm for this proposal! we're excited to review this in more depth.
what this means for service authors is that they would have a standard
system component and format for to track custom billing metrics for service
usage other than "how long was this service deployed". consider an email
service, you could track usage for every email sent or for a data service
you could track the amount of disk usage for data storage. it opens up a
lot of flexibility for how to do chargeback and showback.
for operators, in addition to apps, if the services provided on the
platform use this approach or are translated to this, it provides a
standard way to roll-up usage for a tenant across all of their platform
usage and present it nicely aggregated.
looking forward to the feedback,
james
On Tue, Jul 14, 2015 at 8:30 PM, Michael Maximilien <maxim(a)us.ibm.com>
wrote:
Hi,
This email is to introduce and propose project Abacus to the CF community
for incubation.
*What?*
The IBM Bluemix PaaS (based on CloudFoundry) includes a billing and
metering component such that users can project and drill down on the costs
they have incurred or will incur as they use the system.
Part of this component includes not only APIs that service brokers use to
provide usage data, but also, how to aggregate that information, as well as
a component to expose the aggregated data into a format (JSON) that can be
subsequently consumed by users — this is done via a drill down UI in
Bluemix.
Project Abacus is an extracted Open Source Software version (under Apache
2.0 license) of the APIs, usage metering, and aggregation components from
the Bluemix feature.
Finally, it’s important to state that project Abacus uses a micro-services
architecture and is written in Node.js with CouchDB compatible DBs as
datastore and for map/reduce aggregations. Also, all components can
actually run as CF apps or be externalized. In Bluemix they run as apps.
In other words, this is a bit different from other CF components, however,
we have reviewed Abacus with our colleagues from Pivotal and they also
believe it warrants incubation.
*When?*
Now.
Find the first self-contained release here:
https://github.com/jsdelfino/cf-abacus/archive/v0.0.1.tar.gz
There is a comprehensive README
<https://github.com/jsdelfino/cf-abacus/blob/master/README.md>that can
help you setup and try this first release. The team has included a fake
broker that can be used to generate usage data so you can try the whole
thing out of the box.
*How?*
We plan to introduce this as part of the CAB call discussion tomorrow
07/15 and field any questions.
The next step will be to get approval and become an incubation project.
Once this is done we will be staffing the project with a team (many from
the Bluemix engineers that created it) as well as a PM, and follow the CFF
guidelines for new projects.
We are ready to move on this with no delays and we welcome all interested
parties to try out the first release and to help us refine the APIs and
subsequent releases.
------
Best,
IBM CF team
dr.max
ibm cloud labs
silicon valley, ca
Sent from my iPhone
_______________________________________________
cf-dev mailing list
cf-dev(a)lists.cloudfoundry.org
https://lists.cloudfoundry.org/mailman/listinfo/cf-dev
--
Thank you,
James Bayer
