Making your data accessible to ArcGIS Server

ArcGIS Server makes your data, such as maps, tools, imagery, locators, and so on, available to many other computers in your organization and potentially the entire Internet. However, your GIS server machines need to be able to freely access your data in order to expose it effectively. There are three main things you need to do to make your data accessible to ArcGIS Server:

This topic covers all of the above requirements. First, it's helpful to understand how ArcGIS Server stores and references the data behind its services.

How ArcGIS Server stores and accesses data

When you publish an item as a service, that item is placed on the server. Thus, if you publish a locator as a geocode service, a copy of the locator ends up on the server where ArcGIS Server has sufficient permissions to access it.

Sometimes, you might publish an item that references additional data. For example, a map might reference various feature classes stored in a database. This data is only moved to the server if you do not register the data with the server when publishing. Alternatively, if you do not want the data to be copied to the server at the time of publishing, you need to explicitly take the steps below to ensure ArcGIS Server can see and access your data.

Storing your data where all GIS server machines can see it

Each GIS server machine in your deployment needs to be able to read your GIS resources and all the data they reference. For example, when you publish a map as a service, all the data for the map's layers must be visible to all GIS server machines. Following are some approaches for how you can store the data.

Storing the data locally on each GIS server machine

When you save your data to a local path, for example, /opt/local/data and create a service from it, other GIS server machines are not able to work with the service unless they have their own copies of the data residing at /opt/local/data. Loading an identical copy of your data into an identical path on each GIS server machine can be beneficial for performance, but it may not be a practical solution for large or frequently changing datasets.

Storing the data in a shared directory

Another way to make your data available to all GIS server machines is to use the operating system tools to share the directory in which the data is stored. Shared network directories are commonly referred to with NFS mounted folders, which contain the name of the server (for example, /net/myserver/opt/local/data). When you use NFS paths to reference your data, all GIS server machines will look to the correct machine for the data.

If you store your GIS resources in shared directories, remember that all data source paths within the resource must also use NFS paths or relative paths. For example, if your map document contains layers from three feature classes, the paths to those feature classes must be NFS or relative paths.

Although shared network folders are convenient for referencing data, they require network traffic and can introduce performance bottlenecks that would not otherwise exist when accessing the data through local paths.

Storing the data in a database

Many GIS shops store large data collections in a relational database management system (RDBMS) such as DB2, Informix, Netezza, Oracle, or PostgreSQL. Esri natively supports these databases, which provides a way to conveniently organize and use spatial data from an RDBMS in ArcGIS.

Esri also provides a geodatabase that you can deploy within your RDBMS. Geodatabases and databases are convenient and powerful, but, because they are accessed over the network, they often don't yield as fast performance as you would see when accessing the data locally. In some cases, you may be able to use the geodatabase to replicate to other formats that are easy to store locally, such as file geodatabases. You can then place the data on each GIS server machine for optimum performance.

See Data storage considerations for an ArcGIS Server site to learn about whether this approach is right for you.

Grant the ArcGIS Server account permissions to your data

When you log in to your own computer, the account name you use gives you access to all your files and folders on the computer. No one else can access your data unless you allow them to. The same holds true for your GIS data. The ArcGIS Server account needs to have at least read permissions to any data in folders that you use in your services, and any data in databases that you access using operating system authentication. In some scenarios where edits are occurring, the ArcGIS Server account may also need to have write permissions.

When do you need to apply permissions?

The items you publish are copied to the server where the ArcGIS Server account has already been granted permissions. However, the data referenced in those items (for example, the layers in a map or globe) may or may not have the correct permissions applied, depending on whether you choose to register the containing folder or database with the server when publishing.

If you chose to have the server automatically copy data to the server, there is no need to set any additional permissions. All the data is copied to the server where the ArcGIS Server account already has permissions. If your source data is file based and you are publishing a feature or transaction-enabled WFS (WFS-T) service, the source data is copied into a geodatabase that you have registered with the server, called ArcGIS Server's Managed Database. You do not need to grant permissions to the ArcGIS Server account to access the source data that is copied to the server.

To learn more, see Copying data to the server automatically when publishing.

If you chose to register the containing folder, you need to explicitly give the ArcGIS Server account permissions to read from that folder. For example, when you publish a globe service, you don't need to give the ArcGIS Server account permissions to read your 3DD file (because this is always copied to the server). However, you need to explicitly give the ArcGIS Server account permissions to the data referenced by the layers of that 3DD document.

If you choose to register the containing database, the type of permissions you need to grant depends on what type of database you are using and what type of authentication you are using to connect.

The process of granting permissions to your file-based or database data is described in the remaining sections of this topic.

Granting permissions for file-based data

If your data is file based, such as shapefiles, image files, and file geodatabases, you'll need to work with the operating system to set access to the folders that contain your data. The ArcGIS Server account must have at least read access to the data and write access if the data will be edited. Here are some scenarios:

Granting permissions to data in a database

When you create a service that references data in a database, you need to ensure that the server has the appropriate permissions to access the data. The type of permissions you need to grant depends on what type of database you are using and what type of authentication you are using to connect.

NoteNote:

ArcGIS 10.1 for Server and later releases do not support personal geodatabases.

The way you grant access to data in a database depends on whether you connect to the database using database authentication or operating system (OS) authentication. How can you tell which type of authentication is being used? If the data is in a Desktop or Workgroup geodatabase (accessed through the Database Servers node in ArcGIS for Desktop), OS authentication is used. If the data is in an enterprise RDBMS (Oracle, DB2, Informix, Netezza, PostgreSQL, or Teradata), you can view the connection properties in ArcCatalog or the Catalog window in ArcMap to find out whether it uses database authentication or OS authentication.

NoteNote:

Joins that use ODBC connections on Windows are not supported by ArcGIS for Server (Linux).

Using database authentication

When using database authentication, check your database connection properties in the Catalog tree and make sure you've checked the option to save the user name and password. If you create a map or globe document that uses data through that connection, the name and password will be saved in the map or globe document, and your service should be able to access the data successfully.

Write permissions on the data must be granted to the database user making the connection if you plan to allow edits to the data.

Using OS authentication

If your data will be accessed through OS authentication, you must add the ArcGIS Server account to the database, then grant it permissions to the resources that it needs to access. When the service runs, it will log in to the DBMS as the ArcGIS Server account.

The way that you add the ArcGIS Server account and grant it permissions can vary. You may find it helpful to consult your DBMS documentation to learn how to grant access to an operating system account. Once you've added the ArcGIS Server account, you need to grant it SELECT permissions to the resources that you are going to publish. Write permissions on the data are required if you plan to allow edits to the data.

Registering your data with the server

After you have granted the ArcGIS Server account the appropriate permissions to the folders and databases that contain your data, you need to register them with the server using ArcGIS Server Manager or ArcGIS for Desktop. Data registration gives you the most control over how your server will access data and helps you ensure that the data is truly accessible by the server.

For full instructions, see the following topics:

Copyright © 1995-2014 Esri. All rights reserved.
12/18/2014