MonetDB Docker

Docker container for MonetDB. Dockerfile and scripts can be found in the MonetDB Docker repo.

Usage

The simplest way to start a MonetDB container is to use the images published on Docker Hub.

docker run -P -it monetdb/monetdb:latest

This command will start a container from the latest tagged image on docker hub and it will map a random port on localhost to the port 50000 where the MonetDB daemon is listening for connections (the effect of the -P flag). Alternatively you can specify the port on the local host using the flag -p.

The docker image creates the database farm on the directory /var/monetdb5/dbfarm/ in the container the first time it starts. You can mount a local directory on that path in the container in order to have access to the data from the local host.

The image accepts a number of parameters in the form of environment variables (i.e. passed using the -e docker flag) that will configure the behavior of the container.

• MDB_FARM_DIR

  The location of the database farm. Defaults to /var/monetdb5/dbfarm.

• MDB_DAEMON_PASS

  This is the passphrase used to manage the farm from outside the container. For example:

  monetdb -h localhost -P <passphrase> create SF-1
  

  See also MDB_DAEMON_PASS_FILE, which may be more secure. If neither MDB_DAEMON_PASS nor MDB_DAEMON_PASS_FILE are set, the farm cannot be managed from outside the container

• MDB_DAEMON_PASS_FILE

  Variant of MDB_DAEMON_PASS. Path to a file inside the container that contains the daemon password on the first line. This can be used for example with Docker swarm secrets or Kubernetes secrets management, or manually by mounting a volume that contains the password file.

• MDB_LOGFILE

  The file where the daemon should write the log messages. By default it's the file merovingian.log in the database farm directory in the container.

• MDB_SNAPSHOT_DIR

  This is the directory in the container where database snapshots will be written. If no value is given the daemon will not produce any snapshots. You can mount a local directory in order to have access to the snapshots from the local host.

• MDB_SNAPSHOT_COMPRESSION

  This specifies the compression algorithm to be used for snapshot files. Default value is .tar.lz4, other possible values are .tar, .tar.gz, .tar.xz and .tar.bz2.

• MDB_CREATE_DBS

  This specifies databases to be created the first time the container comes up. You can specify multiple databases by separating their names with a single comma character (no spaces between them):

  [...] -e MDB_CREATE_DBS=db1,db2,db3 [...]
  

  The default is monetdb. If this variable is not empty, MDB_DB_ADMIN_PASS must also be set.

• MDB_DB_ADMIN_PASS

  The password to use for the monetdb (admin) user in the databases created through MDB_CREATE_DBS. Note that all the databases get the same admin password. See also MDB_DB_ADMIN_PASS_FILE, which may be more secure. If neither MDB_DB_ADMIN_PASS nor MDB_DB_ADMIN_PASS_FILE are set, MDB_CREATE_DBS must be set to the empty string.

• MDB_DB_ADMIN_PASS_FILE

  Variant of MDB_DB_ADMIN_PASS. Path to a file inside the container that contains the daemon password on the first line. This can be used for example with Docker swarm secrets or Kubernetes secrets management, or manually by mounting a volume that contains the password file.

• MDB_FARM_PROPERTIES

  A comma separated list of entries key=value to pass to the database farm. This is used in order to set properties like port, exittimeout etc. See monetdbd manual page for the full list of properties. Please note that the properties listenaddr, control and passphrase are not affected by this variable. listenaddr is always set to all, and the other two are set appropriately by setting the variable MDB_DAEMON_PASS or MDB_DAEMON_PASS_FILE.

• MDB_DB_PROPERTIES

  A comma separated list of entries key=value to pass to every database specified in MDB_CREATE_DBS. See the documentation of the set sub-command in the monetdb manual page for the full list of properties.

Note

There are situations where error messages produced at the startup of the container are not immediately visible. For example when using Github Actions, any error messages are shown at the end of the job in the section Stop containers.

Another similar situation is at the command line when the container is created in two steps using the commands docker create and docker start. In that case you can access the error messages using the command docker logs <container name>.

Access

Once the image is running you can get a shell in it:

docker exec -it <image_name> bash

Building the image manually

Clone this git repository and run:

docker build . -t <local tag>

Then you can use the image you just built as described in section Usage above.