AskDot

Data-product creation guide

Each data-product is an independent spark job that runs in a spark-submit mode for generating reports and data migrations. So, even for a new data-product, we will have to add a new scala script with some class base classes extended.

Data-product execution flow

Exhaust job execution flow

As mentioned in Data-product execution overflow, all the data-products are under the JobExecutor from sunbird-core-dataproducts. Thus, before creating a data-product, dependency libraries need to be setup.

Required Baseclasses a new Data-product:

  • IJob

    It is an abstract class in from sunbird-core-dataproducts which used to represent script as data-product job to the job manager

  • BaseReportsJob

    It has the spark utility functions such as creating a spark session for a data-product.

Configurations

Data-product can be executed with following two levels of cofiguration.

application level config

This config is provided from the application.conf file which is common for all the data-products and will not be modified frequently.

Github Path for the template which is used to create the .conf file:

lern-model-config

Each job is collecting data from different data-providers and has various types of inputs. So, model config is implemented to serve data-product level configuration

Github Path for the template which is used to create the model-config file:

Script changes for creating a new data-product

Since the data-product is a batch processing scripts, in server data-products are triggered using shell scripts. So whenever the data-product is implemented we need to add the job id and model-config in the below shell script templates.

lern-run-job : job-id which will be used as identifier and respective data-product classpath will be added in this scripts.

lern-model-config : Respective job id job config has to be added in the this script

PreviousInstallation ConfigurationNextTroubleshooting a data-product

Last updated