The same steps to backup EDG are also applicable for Explorer.
- Stop your EDG server, because the backup must occur offline.
- Backup your entire workspace, by performing a tar or zip operation on it.
- Backup your configured database (TDB, RDBMS). Note: This step depends on what persistence store you have configured, e.g., see your RDBMS manufacturer’s website for more information. NOTE: Be sure to backup the correct database/schema, as configured for your EDG server.
- Store these two files together offline.
- Start EDG backup once all of the above steps have been completed.
- Stop EDG server.
- Locate and delete your current EDG workspace. The workspace location is stored in the web.xml file.
- Replace the existing EDG workspace with your backed-up copy.
- Restore your database from the backup.
- Before starting EDG confirm that your EDG is set up to point at your newly restored database. This can be done by editing [Workspace]/server.topbraidlive.org/dynamic/config.ttl and confirming settings, e.g., the URL points at the newly restored database.
- Start EDG server.
Backup data stored in TDB databases
It is possible to take a filesystem copy of these databases if the EDG server is stopped. If the server is running, the database snapshot may be invalid.
In addition to the backup procedures describes above for a complete server backup, it is possible to backup server-held data for data held in shared TDB-backed storage and the Data Platform EDG cache database can be backed up during normal operation of the EDG server.
NOTE: This only creates a back up of the data stored in <workspace>/_Data/TDB/, the connector (.xdb) files will need to be backed up separately.
This is provided with a web API for scripts to invoke.
The backup service is available to administrators at:
and is accessed with an HTTP POST request (and not an HTTP GET request).
curl -v -XPOST 'http://localhost:8080/tbl/backup?storage=xdb' --output test.trig
Example: to capture all graphs held in the database for the Data
curl -v -XPOST 'http://localhost:8080/tbl/backup?storage=dp'
In each case, the result is a RDF file in standard RDF TriG format with addition information for restoring per-graph prefixes if required.
- Stop the EDG server.
- Make sure you have Jena 3.7.0 installed and added to your PATH (OSX/Linux) or your Environment Variables (Windows).
- Use tdbloader from the Apache Jena TDB tools to build a TDB database (example below).
- Replace the database in the EDG deployment.
- Restart the server.
Example: to build a new database with the exported trig file above:
tdbloader --loc /your/database/location/_Data/TDB test.trig
NOTE: This just restores the database, the connector (.xdb) files should be restored separately into the relevant project folder (Repositories) location.
The database locations are:
|Shared graph TDB storage||
|Data Platfrom EDG cache storage||
Per-graph prefixes can be restored by creating an RDF Turtle with just the prefixes loading that file into the graph.
Backup and Recovery of Supported Relational Databases (RDBMS)
|Database||Backup and Recovery|
|MS SQL Server||https://msdn.microsoft.com/en-us/library/ms187510.aspx|