Install Genomics DB Infrastructure using Ansible - GenomicsDB/GenomicsSampleAPIs GitHub Wiki
Ansible Playbook Description
The Ansible playbook - genomicsdb will setup the following
- Users and Group for genomics db
The playbook provides the option of setting up- Owner User - who will have all the privileges to the directories, data and the postgresql DB
- Client User - who will have read-only access to the directories, data and the postgresql DB
- Directories for data and repos
- Pulls the git repos
- Compiles the search_library
- Creates the postgresql database with the right permissions/privileges, and creates the tables
- Creates the python virtual environment, and installs all the relevant packages under requirements.txt
- Updates the bashrc of the genomics db users with the environment variables required for genomics db
The variables listed below can be overridden (if necessary). The defaults can be found in defaults/main.yml under the genomicsdb role.
| Name | Description |
|---|---|
| meta_db_name | Name of the postgresql database for the Meta database in genomics db |
| metadb_adapter | postgresql+psycopg2 is required by the current implementation. |
| metadb_uri | URI where the Postgresql database is available |
| group_name | Group name for the ga4gh users |
| client_user_name | User name for the client user |
| owner_user_name | User name for the owner user |
| permissions | Permissions to apply |
| genomicsdb_dir | This is the base path where the repos and workspace directories are created. |
| workspace_name | Name of the workspace where Tile DB data will be stored |
| ssh_proxy | In some network configurations, ssh to github may require say a ProxyCommand. If it is applicable to your network, this variable is used. |
| user_libs | This is a array variable that defines the list of additional packages that needs to be installed. This is intended to be used in conjunction with ssh_proxy. Say if ProxyCommnad uses ncat, user_libs can be used to request installation of nmap package |
| mpi_path | Path to where the openmpi package is installed. |
| mpi_runner | Path to where the mpirun executable is located. This differs by default for various operating systems. |
| is_clean_compile | true/false that determines if make clean will be run or not before running make |
| store_branch | Branch of store repo to compile on |
| library_build | can be release or debug to compile GenomicsDB |
| uwsgi_socket_path | Path where the unix sockets are created for communication between the nginx and uwsgi services |
| system_services_extension | string to be used for the service extension. E.g. on RedHat family of system "service" is used |