# Databases

## Daily RDFs

RDF stands for *Reporting Database File.* Exivity uses these to store usage data and configuration information.

A *daily RDF* stores the usage data from which Edify produces report data. Any given RDF contains a single DSET, along with internally managed metadata such as column types and prepared report data.

An RDF is created using the [finish](https://olddocs.exivity.io/3.5.4/data-pipelines/transform/language/finish) statement in a [Transformer ](https://olddocs.exivity.io/3.5.4/data-pipelines/transform)task. For any given DSET, there can be a single RDF per day, although there may be many RDF files per day in total (one RDF for each DSET).

RDFs are named according to the [datadate](https://olddocs.exivity.io/3.5.4/advanced/digging-deeper/transformer-datadate) and ID of the DSET they contain, and have a `.rdf` extension. For example a DSET with an ID of `Azure.usage` for the first day of the month will result in an RDF called `01_Azure.usage.rdf`.

An RDF containing usage data is located at `<home dir>/report/<yyyy>/<mm>/<dd>_<dset>.rdf` where:

* `<yyyy>` is the year of the datadate
* `<MM>` is the month of the datadate
* `<dd>` is the day of the datadate
* `<dset>` is the DSET ID that was used to populate the usage data

## The global database

The *global database* uses PostgreSQL (user data is stored at`<home_dir>/system/pgdata`) and contains system-wide configuration data including (but not necessarily limited to):

* Service definitions
* Service rate revisions
* Service rate adjustments
* Service categories
* User accounts
* Report definitions (and related metadata such as sychronisation timestamps)
* Metadata about RDF files created by Transcript
* Usergroups
* Security settings
* Account information
* Job Schedules

{% hint style="danger" %}
The Global database should never be manually modified unless this process is performed by, or under the guidance of, Exivity support staff. It is critical to make a backup copy of this file before any changes are made.
{% endhint %}