For Data Tank 2.0, settings can be self-managed in the Management UI, which is different from Data Tank where settings are configured by Treasure Data.
Data Tank 2.0 requires Access Keys to access the database, unlike Data Tank which provides default users (tank_user and tank_integration_user). You need to create Access Keys in the Management UI.
Learn more about Managing Access Keys.
In Data Tank 2.0, you can create multiple user-defined schemas in the Management UI, unlike Data Tank which provides only one schema called public.
Learn more about Creating a New Schema.
A database user (Access Key) cannot operate on objects in a schema until permission settings, called Access Policies, are applied. You can grant permissions to an Access Key for a specific schema in the Management UI.
Learn more about Managing Access Policies.
Data Tank 2.0 only allows access from specific IP addresses. Add your IP addresses to the IP Whitelist of Data Tank 2.0 in the Management UI.
g:treasuredata-main is set by default and is an alias for the IP addresses used by each Treasure Data function (including the Management UI). Do not delete it, as doing so will make Data Tank 2.0 inaccessible.
Learn more about Configuring your IP Whitelist.
No data is stored when Data Tank 2.0 is initially provisioned. Prepare any necessary data for import.
If you want to use records stored in Treasure Data tables, use Result Export to PostgreSQL. This feature exports the results of Hive/Presto queries to PostgreSQL.
If you want to move records stored in Data Tank to Data Tank 2.0, follow these steps:
- Import your data from Data Tank into a Treasure Data table using Import from PostgreSQL.
- Extract the imported records with a Hive/Presto query and export the results into Data Tank 2.0 with Result Export to PostgreSQL.
This must be done on a table-by-table basis.
Exporting results to PostgreSQL only exports records. If you need additional objects such as indexes, you need to create them separately. You can use third-party tools such as the psql client or pgAdmin.
There are several methods of integrating with Data Tank. If you replace Data Tank with Data Tank 2.0, you need to update the integrations to change the destination. Examples of such scenarios include:
- Integrations importing from Data Tank
- Integrations exporting to Data Tank
- Running queries via the
pg>operator in Workflow - Third-party tools integrated with Data Tank, such as Tableau
Learn more about changing integrations from Data Tank to Data Tank 2.0.