Virtuoso docker

Docker for hosting Virtuoso.

The Virtuoso is built from a specific commit SHA in https://github.com/openlink/virtuoso-opensource. This image is currently build from commit f3d88f16bca4274265160e098be3ba3c7d68341c, which corresponds to virtuoso 7.2.10. You can build this image from a different commit by providing the correct commit id as the VIRTUOSO_COMMIT build argument.

Running your Virtuoso

docker run --name my-virtuoso \
    -p 8890:8890 -p 1111:1111 \
    -e DBA_PASSWORD=myDbaPassword \
    -e SPARQL_UPDATE=true \
    -e DEFAULT_GRAPH=http://www.example.com/my-graph \
    -v /my/path/to/the/virtuoso/db:/data \
    -d redpencil/virtuoso

The Virtuoso database folder is mounted in /data.

The Docker image exposes port 8890 and 1111.

Docker compose

The image can also be configured and used via docker-compose.

db:
  image: redpencil/virtuoso:1.0.0
  environment:
    SPARQL_UPDATE: "true"
    DEFAULT_GRAPH: "http://www.example.com/my-graph"
  volumes:
    - ./data/virtuoso:/data
  ports:
    - "8890:8890"

Upgrading

There are multiple ways of upgrading your virtuoso version. The procedure described here takes a bit longer, but will result in using all of the latest features of your new virtuoso version and optimizes your DB size on disk.

NOTE: Upgrading virtuoso is a procedure to be done with great care, make sure to have backups before starting.

1. dump nquads

When upgrading it's recommended (and sometimes required!) to first dump to quads using the dump_nquads procedure:

docker compose exec virtuoso isql-v
SQL> dump_nquads ('dumps', 1, 1000000000, 1);

2. stop the db

docker compose stop virtuoso

3. remove old db and related files

When this has completed move the dumps folder to the toLoad folder. Make sure to remove the following files:

• .data_loaded
• .dba_pwd_set
• virtuoso.db
• virtuoso.trx
• virtuoso.pxa
• virtuoso-temp.db

mv data/db/dumps/* data/db/toLoad
rm data/db/virtuoso.{db,trx,pxa} data/db/virtuoso-temp.db data/db/.data_loaded data/db/.dba_pwd_set

Consider truncating or removing the virtuoso.log file as well.

4. update virtuoso version

Modify the docker-compose file to update the virtuoso version

   virtuoso:
-    image: redpencil/virtuoso:1.0.0
+    image: redpencil/virtuoso:1.2.0-rc.1

5. start the db

Start the DB and monitor the logs, importing the nquads might take a long time .

docker compose up -d virtuoso
docker compose logs -f virtuoso

After that your application can be started again and you should be good to go.

Configuration

dba password

The dba password can be set at container start up via the DBA_PASSWORD environment variable. If not set, the default dba password will be used.

SPARQL update permission

The SPARQL_UPDATE permission on the SPARQL endpoint can be granted by setting the SPARQL_UPDATE environment variable to true.

CORS

You may want to enable basic CORS headers on the SPARQL endpoint, this can be done by setting the ENABLE_CORS environment variable to any value. If not set (the default), no cors headers are sent.

.ini configuration

All properties defined in virtuoso.ini can be configured via the environment variables. The environment variable should be prefixed with VIRT_ and have a format like VIRT_$SECTION_$KEY. $SECTION and $KEY are case sensitive. They should be CamelCased as in virtuoso.ini. E.g. property ErrorLogFile in the Database section should be configured as VIRT_Database_ErrorLogFile=error.log.

Dumping your Virtuoso data as quads

Enter the Virtuoso docker, open ISQL and execute the dump_nquads procedure. The dump will be available in /my/path/to/the/virtuoso/db/dumps.

docker exec -it my-virtuoso bash
isql-v -U dba -P $DBA_PASSWORD
SQL> dump_nquads ('dumps', 1, 10000000, 1);

For more information, see http://virtuoso.openlinksw.com/dataspace/doc/dav/wiki/Main/VirtRDFDumpNQuad

Loading quads in Virtuoso

Manually

Make the quad .nq files available in /my/path/to/the/virtuoso/db/dumps. The quad files might be compressed. Enter the Virtuoso docker, open ISQL, register and run the load.

docker exec -it my-virtuoso bash
isql-v -U dba -P $DBA_PASSWORD
SQL> ld_dir('dumps', '*.nq', 'http://foo.bar');
SQL> rdf_loader_run();
SQL> checkpoint;
SQL> checkpoint_interval(N);
SQL> scheduler_interval(M);

Note: N and M should be fetched from your virtuoso.ini config by looking for CheckpointInterval and SchedulerInterval respectively.

Validate the ll_state of the load. If ll_state is 2, the load completed.

select * from DB.DBA.load_list;

For more information, see http://virtuoso.openlinksw.com/dataspace/doc/dav/wiki/Main/VirtBulkRDFLoader

Automatically

By default, any data that is put in the toLoad directory in the Virtuoso database folder (/my/path/to/the/virtuoso/db/toLoad) is automatically loaded into Virtuoso on the first startup of the Docker container. The default graph is set by the DEFAULT_GRAPH environment variable, which defaults to http://localhost:8890/DAV.

Creating a backup

A virtuoso backup can be created by executing the appropriate commands via the ISQL interface.

docker exec -i virtuoso_container mkdir -p backups
docker exec -i virtuoso_container isql-v <<EOF
    exec('checkpoint');
    backup_context_clear();
    backup_online('backup_',30000,0,vector('backups'));
    exit;

Restoring a backup

To restore a backup, stop the running container and restore the database using a new container.

docker run --rm -it -v path-to-your-database:/data redpencil/virtuoso virtuoso-t +restore-backup backups/backup_ +configfile /data/virtuoso.ini

The new container will exit once the backup has been restored, you can then restart the original db container.

It is also possible to restore a backup placed in /data/backups using a environment variable. Using this approach the backup is loaded automatically on startup and it is not required to run a separate container.

docker run --name my-virtuoso \
            -p 8890:8890 \
            -p 1111:1111 \
            -e DBA_PASSWORD=dba \
            -e SPARQL_UPDATE=true \
            -e BACKUP_PREFIX=backup_ \_
            -v path-to-your-database:/data \
            -d redpencil/virtuoso

Contributing

Contributions to this repository are welcome, please create a pull request on the master branch.

New features will be tested on redpencil/virtuoso:latest first. Once the image is verified, version branches will be rebased on master.