Data Ingestion

For this exercise, we have uploaded sample_crm.csv file to Amazon S3 bucket. Follow the steps below to select the file for ingestion.



4. Select summit_hol_crm_data.csv and select Delimited from Data format. Sample records from the csv will be displayed. Click Next.



5. Click Existing dataset and select the dataset you created in previous exercise . Set Dataflow name to [your initials] - CRM Profile Dataflow. Click Next.



6. Fields from csv file are automatically mapped to the selected schema using ML algorithms, however it may not always get it right. For example the phone is incorrect here.



Check your mapping and if you see anything that doesn't look right, select the correct mapping by clicking on the Target Field and selecting the right option from the Target Schema on the right.



7. Let's correct the mapping for “city”, Click on the Target Field and select homeAddress.city instead



Click Next once you're happy the fields are correctly mapped.



7. Let's also add the current timestamp as a last modified date to the ingestion. Click on the New field type button and select Add calculated field.



Add the function now() to the formula and show the preview



Click Save. Map the new field to _repo.modifyDate.



8. The next step is to set up the scheduling. Batch based source connectors can check for new data as frequently as every 15 mins, for this exercise however we'll just set it to run once. Leave the start time as is and click Next.



9. Review and click Finish to create your dataflow.



10. The newly configured dataflow will be displayed under the Amazon S3 source connector.





The dataflow will start asynchronously according to the schedule but may take a minute or two to kick off the first execution.