Several configuration options are available, for AskOmics itself, and its various related containers (virtuoso, isqlapi, etc...)

Askomics configuration

There are two type of configuration options in Askomics:

  • Options relevant to the web server itself (number of workers, timeouts, etc...)
  • Options regarding the AskOmics application itself

Please see here for an env file example with both type of variables.

Web server configuration

The following environment variable are available to configure the web server:

  • NTASKS: Maximum number of Celery (background: integration, queries...) tasks in parallel (default to 1).
  • WORKERS: Maximum number of web server workers (default to 1).
  • TIMEOUT: Web server worker timeout, in seconds (default to 300).

AskOmics application configuration

All AskOmics configuration is set in config/askomics.ini files. When AskOmics is deployed with docker-compose, configuration is set with environment variables. The environment variable should be prefixed with ASKO_ and have a format like ASKO_$SECTION_$KEY. $SECTION and $KEY are case sensitive. E.g. property footer_message in the askomics section should be configured as ASKO_askomics_footer_message=Welcome to my AskOmics!

  • flask

    • secret_key (string): AskOmics secret key for session.
    • session_timeout (int): timeout for the session cookie (in minutes)

  • celery

    • broker_url (url): Redis or RabitMQ url
    • result_backend (url): Redis or RabitMQ url

  • askomics

    • debug (true or false): Display debug log in console
    • debug_ttl (true or false): Keep converted rdf file. Set the to true can fill you disk, active only for debug purpose
    • reverse_proxy_path (string): Proxy path if AskOmics is accessible under a subpath
    • subtitle (string): Subtitle, displayed on the browser tab
    • footer_message (string): Custom message displayed on the AskOmics footer
    • front_message (string): Custom message displayed on the AskOmics main page
    • contact_message (string): Custom message displayed on the Contact page
    • display_commit_hash (true or false): diplay the commit hash of the current version in the AskOmics footer
    • data_directory (path): Where AskOmics store the data
    • database_path (path): Path to the sqlite database
    • disable_account_creation (true or false): disable the possibility to create new accounts
    • npreview (int): Number of line displayed during integration
    • password_salt (string): Password salt
    • disable_account_creation (true or false): Disable new account creation
    • default_locked_account (true or false): Lock new account
    • disable_integration (true or false): Disable integration to non admin users
    • protect_public (true or false): Public datasets and queries are visible only for logged users
    • enable_sparql_console(true or false): Allow non-admin logged users to use the sparql console. This is unsafe.
    • quota (size): Default quota for new users (can be customized individually later)
    • github (url): Github repository URL
    • instance_url (url): Instance URL. Used to send link by email when user reset his password
    • smtp_host (url): SMTP host url
    • smtp_port (int): SMTP port
    • smtp_user (string): SMTP user
    • smtp_sender (email): SMTP sender
    • smtp_password (string): SMTP password
    • smtp_connection (string): SMTP connectin (starttls or null)
    • ldap_auth (true or false): Use LDAP authentication
    • ldap_host (string):LDAP host
    • ldap_port (int): LDAP port
    • ldap_bind_dn (string): LDAP bind DN string
    • ldap_bind_password (strig): LDAP password
    • ldap_search_base (string): LDAP search base
    • ldap_user_filter (string): LDAP user filter
    • ldap_username_attribute (string): Username attribute
    • ldap_first_name_attribute (string): First name attribute
    • ldap_surname_attribute (string): Surname attribute
    • ldap_mail_attribute (string): Mail attribute
    • ldap_password_reset_link (url): Link to manage the LDAP password
    • ldap_account_link (url): Link to the LDAP account manager
    • autocomplete_max_results (int): Max results queries by autocompletion
    • single_tenant (bool): Enable single tenant mode
    • anonymous_query (bool): Enable anonymous query mode
    • anonymous_query_cleanup (int): Number of days before a successful anonymous job is deleted

  • virtuoso

    • triplestore (string): Triplestore used. Can be virtuoso, fuseki or corese
    • endpoint (url): Triplestore endpoint url
    • updatepoint (url): Triplestore updatepoint url
    • isqlapi (url): isql-api url when AskOmics use isql-api to perform queries on virtuoso through isql. If not set, AskOmics will launch queries on the SPARQL endpoint
    • fuseki_upload_url (url): If triplestore is fuseki, set the fuseki upload url
    • username (string): Triplestore credential: username
    • password (string): Triplestore credential: password
    • load_url (url): AskOmics url accessible from the triplestore
    • upload_method (string): upload method fir virtuoso. Can be load or insert
    • chunk_size (int): Number of RDF triples to upload in one time
    • block_size (int): Size of location bocksize for positionable entities
    • serialization_format (string): RDF serialization format. Can be nt, turtle or xml
    • default_graph (string): Triplestore default graph
    • users_graph (string): User base graph
    • namespace_data (url): AskOmics namespace for data. Correspond to the : prefix. You should change this to your instance url if you want your URIs to be resolved.
    • namespace_internal (url): AskOmics namespace for internal triples. Correspond to the askomics: prefix. You should change this to your instance url if you want your URIs to be resolved.
    • preview_limit (int): Number of line to be previewed in the results page
    • result_set_max_rows (int): Triplestore max row. Must be the same as SPARQL[ResultSetMaxRows] in virtuoso.ini config

  • federation

    • query_engine (string): Query engine to use, can be corese or fedx
    • endpoint (url): Federated query engine endpoint url
    • local_endpoint (url): Triplestore url, accessible from the federated query engine

  • sentry

    • server_dsn (url): Sentry url for the server
    • frontend_dsn (url): Sentry url for the frontend

Virtuoso configuration

Several environment variable are available to configure the Virtuoso instance. You can find an example of a configuration file here

The following parameters are related to virtuoso's performance. Please consult the documentation for more information.

  • VIRT_Parameters_NumberOfBuffers
  • VIRT_Parameters_MaxDirtyBuffers
  • VIRT_Parameters_TN_MAX_memory
  • VIRT_Parameters_TransactionAfterImageLimit
  • VIRT_SPARQL_ResultSetMaxRows
  • VIRT_SPARQL_MaxDataSourceSize
  • VIRT_SPARQL_MaxQueryCostEstimationTime
  • VIRT_SPARQL_MaxQueryExecutionTime

The following are also available:

  • DBA_PASSWORD: Default 'dba' user password. Make sure it matches the AskOmics and isqlapi parameters
  • SPARQL_UPDATE: Allow any user to make 'load' queries. Should be set to False for security reasons.
  • DEFAULT_GRAPH: Default graph location

Isqlapi configuration

Isqlapi is a web application that helps speeding up data loading and querying. Several parameters are available for configuration:

  • WORKER_NUMBER: Number of workers for the web service
  • WORKER_TIMEOUT: Worker timeout for queries
  • ISQL_API_VIRTUOSO_DSN: Virtuoso's dsn
  • VIRTUOSO_HOST: Virtuoso's instance host name
  • VIRTUOSO_ISQL_PORT: Virtuoso's port for the ISQL protocol
  • ISQL_API_VIRTUOSO_USERNAME: Username for the queries
  • ISQL_API_VIRTUOSO_PASSWORD: Password for the queries