Hello Thomas,

Storing data frames in SQL is only supported by the v1 storage interface: https://docs.openstack.org/cloudkitty/latest/admin/configuration/storage.html

The v2 storage interface introduced InfluxDB and Elasticsearch as replacement.

I believe the plans of the previous core team was to deprecate v1 storage [1], so I am not sure we'll want to put lots of effort in supporting it.

However feel free to propose a patch, if it's trivial enough I don't see why we wouldn't merge it.

[1] https://docs.openstack.org/releasenotes/cloudkitty/stein.html#new-features

On Fri, 23 Jul 2021 at 16:25, Thomas Goirand <zigo@debian.org> wrote:

Hi,

Very fast, with only a few resources, the rated_data_frames tables
becomes huge.

I was very surprised to see there's absolutely zero index in this table.
Isn't a "openstack rating dataframes get" querying this table, with
tenant_id and begin / end as filters? Wouldn't the API calls be a lot
more efficient with such indexes?

Cheers,

Thomas Goirand (zigo)