Table of Contents

• What is the VLO?
• What should I read?
• Setting up
  • Preparing for release
• Using the VLO
  • Running an import

What is the VLO?

Using the VLO faceted browser, you can browse metadata by facet. It consists of three software components: a Solr server with VLO specific configuration, an importer and a web application front end.

For more information, see the wiki page.

What should I read?

• README This file; a general introduction and development information
• DEPLOY-README Instructions on how to deploy a fresh VLO installation
• UPGRADE Instructions on how to upgrade an existing VLO installation
• CHANGES A list of changes per release
• LICENSE Licensing information

Licensing

All code unless otherwise noted is licensed under the terms of the GNU GENERAL PUBLIC LICENSE 3 (GPL3). Please refer to the file LICENSE in the root directory of this repository.

Development information

Some general development notes.

Setting up

Follow the instructions in DEPLOY-README to set up a development environment. You may deploy the Solr instance and the web app from your IDE. Make sure to set the required Java system property for the Solr data location (solr.data.dir)!

Preparing for release

These instructions apply to any kind of release, whether it's a stable version, or for beta or alpha deployment. Always increase the version number and keep the trunk in -SNAPSHOT but not the tags so that a deployed version can always be traced back to its sources!

• Make a tag of the version to release:

  ` git checkout development git checkbout -b release-vlo-4.a.b

• Change the version number in the poms to match the release (should match the directory name and be non-snapshot!!):

  mvn versions:set -DnewVersion=4.a.b

This will update the version numbers of the parent pom and all VLO modules in one go!

• Build the tag and inspect the output of vlo-distribution

  mvn clean install #do not skip unit tests ;)

Unpack the tarball in vlo-distribution/target somehwere and check its contents on version numbers, config files etc.

You may also want to do a 'svn diff' to check the change of the version numbers.

• Clean up and commit

  mvn versions:commit #cleans up POM backups mvn clean #cleans up build outpit svn commit -m "Created tag for VLO version 4.a.b" git commit

• Merge into master and tag to finalise the release

  ``` git checkout master git merge release-vlo-4.a.b #maybe some conflicts needs to be resolved after this git push

  tag

  git tag -a 4.a.b git push --tags ```

• Merge changes into development

  git checkout development git merge release-vlo-4.a.b #maybe some conflicts needs to be resolved after this mvn version:set -DnewVersion 4.c-SNAPSHOT mvn versions:commit git commit git push

• Done!

After building the entire project, a deployment package will be present in the 'target' directory of vlo-distribution. This includes WARs for both the Solr and the web app front end as well as the importer script and default configuration files.

Running the VLO

After a successful deployment and configuration, you should be able to browse to the VLO web app and browse the imported records.

If the VLO is empty, you will need to run an import first.

Running an import

To run an import, go the 'bin' child directory of the VLO application directory and run

./vlo_solr_importer.sh

as the appropriate user (e.g. 'vlouser').

It's advisable to run this in a detached background process (for example using "screen") because an import can take quite a long time depending on the amount of records to be imported. Also make sure that enough memory is available. Some VM parameters are configured inside the script.

Some progress information is logged to a file 'log/vlo-importer.log'. It gets rotated automatically by the import process.

For a fresh import, even when the VLO is not configured to delete all documents from the index first, you can simply remove the contents of the Solr data directory (check server configuration documentation for the exact location).