If you use the ArcGIS for Server (Ubuntu Linux) Amazon Machine Image (AMI) to create an ArcGIS Server site, you have the option to include a PostgreSQL database cluster. The database cluster can be on the ArcGIS for Server instance or on a separate instance that is part of your ArcGIS for Server site. If you use the ArcGIS for Server (Ubuntu Linux or Windows) AMI, you have the option to include an Amazon Relational Database Service (RDS) for PostgreSQL. If you use Amazon RDS for PostgreSQL, the RDS instance is always on a separate instance that is part of your ArcGIS for Server site.
When you use ArcGIS Server Cloud Builder on Amazon Web Services to launch your ArcGIS Server site, the database cluster contains two pre-created geodatabases: egdb and geodata.
The egdb geodatabase is registered as an ArcGIS Server managed database. When you publish feature or WFS-T services to a site that has a registered managed database, the data can be copied from the geodatabase that contains the map source data (publisher geodatabase) to the registered database (server geodatabase). This data is dependent on the service; when you delete the feature or WFS-T service, the data is deleted from the egdb geodatabase.
The geodata geodatabase is intended for use as a replicated geodatabase. You can register the geodata geodatabase with ArcGIS Server. When you do so, designate the geodata geodatabase as a server database connection that is not the same as your publisher database connection (your on-premises enterprise geodatabase) and create a geodata service. You can replicate data from your on-premises enterprise geodatabase to the geodata geodatabase through the geodata service.
When you publish a feature or WFS-T service that includes the replicated data, edits made to the data through the feature service can be synchronized with the geodata service, updating the data in your on-premises enterprise geodatabase. Similarly, you can continue to edit your on-premises data and use the geodata service to synchronize those changes to the data in the geodata geodatabase.
The PostgreSQL database cluster also contains three database login roles:
- postgres: The database cluster administrator
- sde: The geodatabase administrator*
- owner: A user who owns a schema in the egdb and geodata databases and, therefore, can create data in them*
By default, the passwords for these users are as follows:
- postgres = postgres
- sde = E$ri3774*
- owner = owner*
*These login roles are only created if you use ArcGIS Server Cloud Builder on Amazon Web Services to create your site.
To secure your site, you must log in to the instance that contains the PostgreSQL database cluster and change the passwords. See Change default PostgreSQL passwords for instructions.
Amazon RDS for PostgreSQL
You can choose to use an Amazon Relational Database Service (RDS) for PostgreSQL to store your ArcGIS web service data on Amazon.
Amazon RDS performs some database maintenance tasks so you don't have to. For example, RDS instances automatically apply database patches and create backups of your database. They are also designed to be highly available, as they can be run in multiple availability zones (referred to as multi-AZ deployments).
Multi-AZ deployments are the default deployment type for Amazon RDS instances; therefore, Amazon RDS for PostgreSQL instances you launch using ArcGIS Server on Amazon Web Services AMIs use this deployment.
When you use Amazon RDS for PostgreSQL in your ArcGIS Server on AWS site, the Amazon RDS for PostgreSQL instance is always separate from your ArcGIS Server instance.
Geodatabases created in Amazon RDS for PostgreSQL stores spatial data using the PostGIS geometry storage type.
When you use ArcGIS Server Cloud Builder on Amazon Web Services to create your site, it creates the geodatabases and users described previously in this topic. If you manually create your site using the Amazon Web Services console and want to use PostgreSQL RDS for data storage, see Create a geodatabase on Amazon Relational Database Service for PostgreSQL.