English

How To: Migrate from GPT 3.0 to GPT 3.1

Summary

The migration steps outlined in the GPT 3.1 Installation Guide discuss how to migrate a GIS Portal Toolkit (GPT) 3.0 database from GPT 3.0 to GPT 3.1. Instructions provided only refer to migrating the database. They do not cover migration of the Portal and Map Viewer Web applications, because the JavaServer Pages (JSPs) in nearly all sites are uniquely customized and cannot be subject to template-like migration procedures.

Procedure

The steps below outline how to migrate the GPT database and provide tips on the easiest way to integrate customizations made to the GPT interface.

  1. Archive the GPT 3.0 Portal and arcexplorer folders from <Tomcat Installation Dir>\webapps to a backup location.
  2. Delete the existing <Tomcat Installation Dir>\webapps\Portal and <Tomcat Installation Dir>\webapps\arcexplorer folders.
  3. Database Migration Method 1:
    Create a new database instance that is associated with the new 3.1 Portal.

    Database Migration Method 2:
    Another option is to run the migrate_v30v31_oracle.sql or migrate_v30v31_sqlserver.sql database migration scripts.

    For either method, follow the instructions in section "3.0: Windows-Based Portal Database Installation" of the GPT 3.1 Installation Guide.
  4. Now that the database is ready for GPT 3.1, deploy the new WAR files in Tomcat. To do this, copy the new WAR files from the source location to the <Tomcat installation folder>\webapps location and restart Tomcat.
  5. Run the Post Install Configuration wizard, to configure the key properties and settings of the GPT.
  6. If Database Migration Method 2 was chosen in step 3 above, run the migrateV30V31.jsp as explained in section 4.2 of the GPT 3.1 Installation Guide.
  7. All customizations, such as images, styles, or changes to the gpt.properties files must be done manually. The archived GPT 3.0 files are useful because in most cases, customizations can be copied and pasted from the old files into the new files.
  8. If Database Migration Method 1 was chosen in step 3 above, harvest the metadata records from the former database instance to the new one. All data in the tables storing information on user accounts, channels and service permissions (in the TOOLKIT_MAPSERV_PERM table) has to be manually transferred between the databases.