Tool for moving SSP data from one data to another by representing that data as Liquibase changsets.
This tool will not touch your target database at all and interactions with the source database are strictly read-only.
The Liquibase scripts generated by this tool assume your target database already exists and has been initialized by SSP's standard Liquibase scripts.
There are no special provisions for locking the source database nor guaranteeing that it is in a read-only state. So you are probably best served to run this tool while the source application is not serving requests.
- Configure a schema file describing the tables you want to export from the
source database. The project includes a sample for SSP v1.2.0 at
src/main/groovy/resources/BaseConfigDBSchemathat would be suitable for migrating reference data types typically edited during a formal SSP training session.
- Run the tool, configured with that schema file and coordinates for your source database, redirecting stdout to an XML file.
- Take the resulting XML file and execute it against your target database using Liquibase tools. The simplest way to accomplish the latter is with the Liquibase command line.
This is, of course, not the only way to move data from one database to another. PostgresSQL in particular ships with standard export/import tools which should be fully capable of migrating some or all SSP data between databases, thereby obviating this tool. But the situation has historically been more challenging with SqlServer, which has difficulty scripting the SSP database. Direct database-to-database export/import jobs have worked better, but these are often not allowed by network security policies.
At a minimum you'll need a Java JDK install. This process can vary widely from platform to platform, so you're on you're own for that one. This program will theoretically work on JDK 1.5+, but you'll be best served by 1.6+.
Depending on how you plan to run the application you can either install Groovy or Gradle. Gradle ships with an embedded Groovy, so you can technically avoid a standalone Groovy install if you'd like. If you just want to clone this repo and run the code with a minimum of fuss, Gradle is the way to go.
Gradle download. The latest binary distro should be fine. We've most recently tested with 1.4.
Gradle install instructions.
The rest of this doc assumes you've added Gradle's
bin directory to your PATH.
Groovy download. The latest binary distro should be fine. The app has been tested with 1.8.6 and that's the version Gradle will bundle with the binary output.
Groovy install instructions.
The rest of this doc assumes you've added Groovy's
bin directory to your PATH.
Running the Program
The app can be executed in at least three different ways:
Gradle run command. This is the easiest way to run the app if you are tweaking the source code:
%> gradle -q run -PsysProps="<opts>" -PcliArgs="<opts>"
<opts>parsing is naive so spaces in argument values will not work as expected.
Exploded Gradle-assembled application
%> gradle installApp
%> ./build/install/ssp-data-migration/bin/ssp-data-migration <opts>
<opts>splitting is handled by your shell in this case, so whitespace in argument values should work as expected.
Use this execution mechanism if you've received a zip or tar of the application.
Use the JVM_OPTS env var to set JVM properties when executing the app in this way.
Without Gradle. Don't know why you'd want to do this as it is significantly more verbose and you'll need to do more legwork to make your JDBC driver visible:
%> groovy -cp "src/main/groovy/:src/main/resources:/path/to/your/jdbc/driver" \ src/main/groovy/org/jasig/ssp/util/migration/Main.groovy \ <opts>
Output is an XML document on stdout. Redirect as appropriate.
Options all have POSIX "long opt" format, i.e.
The "=" is optional and can be replaced with whitespace. E.g.:
%> gradle -q run -PcliArgs="--db-url=jdbc:postgresql://localhost:5432/ssp" > migration.xml
--db-url [Required] The full URL to the source database
--db-username [Optional] The username for the source database
--db-password [Optional] The password for the source database. NOTE: If this property is not set you will be prompted to enter a source database password upon opening a connection to the source database.
--schema-file [Optional] Specify the location for the DB Schema file - the file that specifies what data to export (default: March 2013 tables from last export job)
--help Prints the usage message and performs no processing