Closed Bug 1729084 Opened 5 years ago Closed 5 years ago

Add a filter `urlbar_clients_daily` to prevent large spikes in data table

Categories

(Data Platform and Tools :: General, defect)

defect

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: teon, Assigned: anicholson)

Details

We've recently landed a new derived table telemetry.urlbar_clients_daily that allows us to have a product-specific view of our Awesome Bar. in a recent exploration, I've noticed that there was some aberrant behavior on August 10, 11 (https://sql.telemetry.mozilla.org/queries/81003/source#200937). This was caused by a particular client that spiked the data. We don't currently have any filtering that's been applied to our Awesome Bar interactions. We would like to apply similar logic that we have for our search counts, which removed/cap these types of recordings.

Assignee: nobody → akomarzewski
Assignee: akomarzewski → anicholson

Fix is here: https://github.com/mozilla/bigquery-etl/pull/2321 that removes recordings over the same threshold as search. Working on a backfill now from 2021-01-01 (beginning of the dataset)

Backfill is done, https://sql.telemetry.mozilla.org/queries/81003/source#200937 no longer has those spikes! Teon could you verify that this is good?*

hi, i'm looking at https://sql.telemetry.mozilla.org/queries/81007/source#200939 but I'm not seeing 1 among the other positions, which is strange.

Flags: needinfo?(anicholson)

Hey Teon, there was a mismatch between the back-filled dataset and the views. This should be fixed now

Flags: needinfo?(anicholson)

cool, this looks good now. thanks!

Status: NEW → RESOLVED
Closed: 5 years ago
Resolution: --- → FIXED
Component: Datasets: General → General
You need to log in before you can comment on or make changes to this bug.