obda-mixer

Setup

The configuration file (configuration.conf in this guide) contains various information like the location of the mappings file, or the desired path for logging. Parameters are specified in the CSV format:

## Test Config
num-runs 1                                              # Number of test runs
num-warmups 1                                           # Number of warm-up runs
timeout 1200                                            # Query Execution Timeout
num-clients 1                                           # Number of clients
rewriting false                                         # Rewriting [true-false]

## Execution Modes
mode java-api                                           # Mode [java-api, web, shell]
java-api-class it.unibz.inf.mixer_ontop.core.MixerOntop # Name of the `Mixer` implementation, `java-api` mode. 
service-url http://sparql-endpoint.sparql/              # Addr. of the SPARQL endpoint, `web` mode.
shell-cmd cmdName.sh                                    # Name of the shell script to execute, `shell` mode.
shell-out filePath.log                                  # Logging of the shell script output, `shell` mode.

## Database Connection Credentials
db-url localhost/npd
db-username user
db-pwd pwd
driver-class com.mysql.jdbc.Driver

## Ontology, mappings, and log info
owl-file resources/npd-v2-ql.owl
mappings-file resources/npd-v2-ql-mysql-ontop1.17.obda
queries-dir resources/Templates
log-file resources/MixerStats.txt

## Advanced Test Config
forced-timeouts 01.rq 02.rq                             # Space-separated list of queries to consider as timeout
forced-timeouts-timeout-value 1200                      # Timeout to be assigned to the queries specified as `forced-timeout`

Legenda:

• num-runs: Number of test runs
• num-warmups: Number of warm-up runs (i.e., runs whose execution times will not be recorded)
• timeout: Maximum execution time allowed to a query, in seconds. A value of zero means no timeout. This parameter works only in owl-api mode, and only with jdbc drivers supporting the timeout feature.
• num-clients: querying the system in parallel. Rewriting and unfolding times are unavailable in multi-client mode.
• rewriting: If query rewriting is enabled. True or False.
• mode: Execution mode. One of: java api mode (java-api), sparql endpoint mode (web), or shell script mode (shell)
• service-url: URL for the SPARQL Endpoint (To be used with –obda=web)
• db-url: URL of the database that the obda-mixer should use for extracting values in order to instantiate the query templates
• db-username: Username for accessing the database
• db-pwd: Password for accessing the database
• mappings-file: Path to the mapping file
• owl-file: Path to the file containing the ontology
• queries-dir: Location of the directory containing the query templates (Templates folder).
• log-file: File where obda-mixer will write the results of the tests (resources/MixerStats.txt)
• driver-class: Database driver class. Database values are used to instantiate the query templates.
• java-api-class: Class for the Mixer implementation. This parameter should be used in combination with the --mode=java-api option.
• shell-cmd: Command-line string for shell execution. This parameter should be used in combination with the --mode=shell option.
• shell-out: Should the output of the shell command be logged? This parameter should used in combination with the --mode=shell option.
• forced-timeouts: It forces the specified space-separated queries to timeout. The timeout value is specified through --timeout-value the option.
• forced-timeouts-timeout-value: Timeout value, in seconds, assigned to forcefully (through the --forced-timeouts option) timeouted queries. Default is 1200.