Product Analytics (PA) has been outputting campaigns monthly metrics on a monthly basis, on the first Friday of the month and outputting metrics into the metrics sheet. @Iflorez is now to transition the running queries for monthly reporting to the Campaigns Product (CP) team. Product Analytics can continue to consult on the data and interpretation as needed.
####Plan
- Inform the team of the change
- Discuss timing
- Review options:
-- Run current JupyterLab notebook in Jupyter Lab
-- Run a version of the notebook as a python executable file
-- Since [[ https://phabricator.wikimedia.org/T337056#9432185 | Jan 24 ]] x1 cluster wikishared dBs are accessible on Superset's SQL lab -- run queries on SQL Lab and output to [[ https://superset.wikimedia.org/superset/dashboard/fea20d2b-bb32-46fa-8221-f6ed2437d1e8/ | dashboard ]] OR run all in one go and download data to csv to import the new row of monthly data into the existing sheet.
-- crontab?
-- ?
- Until we have a set transition date where Claudio is ready to begin outputting metrics, Irene will continue to output the metrics.
Acceptance Criteria
[x] meet &greet for @cmelo to understand the current process and @Iflorez to understand how Claudio might be running/setting up on his end
[x] identify Superset dashboarding with virtual dataset - potential concerns/risks and share with the team
[x] review with KC
[x] ask about superset dashboarding on the [[ https://wikimedia.slack.com/archives/CSV483812/p1752174106591259 | Slack data-engineering channel ]]
[] update queries per [[ https://wikimedia.slack.com/archives/C020V0GN9P1/p1751907205021679?thread_ts=1749506776.888799&cid=C020V0GN9P1 | Slack ]] discussion; confer with @ifried
[] get all the artifacts in place (queries, notebooks etc)
- T374500: [[ https://gitlab.wikimedia.org/iflorez/data-pipelines/-/tree/main/campaigns?ref_type=heads | queries & HQL files on Gitlab ]]
[] group meet to review options
[] engineers do a dry run to understand the process
[] discussions/meetings where helpful.
[] [Potentially] spikes on the campaigns engineering side / devoted time to automate this more.