# Working with the application database
Tonic Structural uses a PostgreSQL database to store its internal state. This database does not store customer data. It only stores data that Structural needs to operate.
## Selecting the schema for the Structural application database
By default, Structural creates the application database tables in the `public` schema.
To use a different schema, set the [environment setting](https://docs.tonic.ai/app/admin/environment-variables-setting) `TONIC_DB_SCHEMA` to the schema name.
Note that you cannot set this setting from the **Environment Settings** tab on **Structural Settings**. You must [set it from a .yaml file](https://docs.tonic.ai/app/environment-variables-setting#environment-setting-config-yaml) and then [restart Structural](https://docs.tonic.ai/app/environment-variables-setting#restarting-after-a-configuration-change).
Also note that if you change the schema, you should do it when you first install Structural.
If you change the setting on a running instance, it deploys the application schema objects into the new schema beside the existing `public` schema, but it does not migrate the data from `public`.
When the application launches after the schema change, it looks like a new instance of Structural. For example, it has none of your user accounts or workspaces. To restore the data, you must manually move it from the `public` schema to the new schema.
## Connecting to the Structural application database
{% hint style="warning" %}
In general, you do not need to connect directly to the Structural application database. These details are provided if needed for advanced troubleshooting.
{% endhint %}
When you initially set up Structural, you provide Structural with connection details to a PostgreSQL database. How to connect depends on where you set up your PostgreSQL database.
### Inside a Docker container
Some customers set up a PostgreSQL database inside the same Docker network as the other Structural containers. In that case, you should ensure that:
* Port 5432 is exposed on the PostgreSQL Docker container.
* Port 5432 is properly mapped.
To do this, add the following section to the PostgreSQL service section of your **docker-compose** file:
```
#This will expose Port 5432 on the docker container and map it to
#Port 5432 on the host machine
ports:
- 5432:5432
```
### Standalone PostgreSQL or cloud installation
If you installed PostgreSQL in a standalone fashion, or you use a cloud service such as Amazon RDS, then make sure that the firewall settings and security groups allow a connection on the appropriate port.
### Making the connection
You can use any PostgreSQL client. The following example uses psql, the PostgreSQL command line client.
```
#Format
#psql -h -p -U -d
#Example
psql -h localhost -p 5432 -U pguser -d allos
```
## Other Structural application database tips and troubleshooting
### How can I reduce the size of the Structural application database?
The Structural application database includes a `ColumnSchema` table that contains the column schema information for your workspace source databases.
By default, when you delete a workspace, Structural does a soft delete. It marks the workspace as deleted, but does not remove the associated rows from the `ColumnSchema` table.
To help prevent the table from growing too large, you can configure Structural to instead do a hard delete, and remove the `ColumnSchema` rows that are associated with a deleted workspace.
To have Structural remove the `ColumnSchema` rows for deleted workspaces, set the [environment setting](https://docs.tonic.ai/app/admin/environment-variables-setting) `TONIC_DELETE_COLUMN_SCHEMA_ON_WORKSPACE_DELETE` to `true`. You can add this setting manually to the **Environment Settings** tab on **Structural Settings**.