1534354 - Validate telemetry on the new data pipeline

Status: RESOLVED FIXED

(Firefox :: New Tab Page, enhancement, P2)

Description

[Nan Jiang [:nanj]]

Now that we have telemetry mirroring enabled in all pre-release channels, we'd like to verify following items:

• Total number of records in the impression table in BQ matches against the total number of records in the session table in RS. Note that on the RS side, we should only fetch sessions from those users who have Top Stories enabled in the beta and nightly channels.
• Total numbers of impressions/clicks/blocks/save_to_pocket match against the impression table in Redshift. Note that we store an array for impressions in the BQ table, whereas those impression arrays are splitted into individual impressions in Redshift.
• Verify that all the columns (both required and optional) are stored in BQ, and the values are expected.

Comment 1

[Nan Jiang [:nanj]]

Have finished up the first round validation, all the results have been recorded in the Tiles/GCP doc.

Will meet with data team to discuss the next steps.

Comment 2

[Nan Jiang [:nanj]]

Filed a bug for gcp-ingestion to track the issues that we found during the data validation.

Comment 3

[Firefox Bug Husbandry Bot]

Reverting spam.

Comment 4

[Nan Jiang [:nanj]]

Close it as Kenny and I have verified the data in release 68. Everything looks good except a minor issue on the backend side, which has already been fixed here.