Page MenuHomePhabricator

Fix content_translation_event user_edit_count_bucket field enums
Closed, ResolvedPublic

Description

There are two errors in the enums listed for the user_edit_count_bucket field of the content_transtion_event schema. See details below:

  • Error: There is an extra space in the '1-4 edits ' enum. This will cause any events sent by the instrumentation that have "1-4 edits" to be rejected. Recommended fix: Remove extra space in the in the '1-4 edits ' enum
  • Error: Anon user events may be rejected by current instrumentation because the user_edit_count_bucket is assigned as null, which is not one of the allowed values. Rcommended fix: Add an explicit 'N/A' to the enum to use with IP editors

Note from @mpopov on how to implement changes in the schema: "Through manual materialization, we can actually fix the typo without bumping the version. Structurally nothing changes so if we delete all the files except current.yaml, update that (without changing the version specified), and run ./node_modules/.bin/jsonschema_tools materialize jsonschema/analytics/mediawiki/talk_page_edit/current.yaml it'll be fine.

Event Timeline

MNeisler triaged this task as Medium priority.Jan 12 2022, 2:30 PM
MNeisler created this task.
MNeisler moved this task from Triage to Current Quarter on the Product-Analytics board.
MNeisler renamed this task from Fix content_translation_event user_edit_count_bucket field to remove extra space in the `1-4 edit` enum to Fix content_translation_event user_edit_count_bucket field enums.Jan 12 2022, 2:37 PM
MNeisler updated the task description. (Show Details)
MNeisler closed this task as Resolved.EditedAug 19 2022, 3:51 PM

This will be addressed next week. Moving to doing

Change 825871 had a related patch set uploaded (by Bearloga; author: Bearloga):

[schemas/event/secondary@master] content_interaction_event: Fix typo and add new val to enum

https://gerrit.wikimedia.org/r/825871

Change 825871 merged by jenkins-bot:

[schemas/event/secondary@master] content_interaction_event: Fix typo and add new val to enum

https://gerrit.wikimedia.org/r/825871

EventGate caches schemas so the fix won't be live until EG is restarted. There is a plan to restart it later this week for a similar reason (to retroactively add geo data to Android's schemas without bumping versions up).

@MNeisler: The service was restarted as part of T316047. At the very least the '1-4 edits ' validation errors should disappear.

The anon user edit count thing would require the instrument to be updated to set 'N/A'. So until then you'll still see those 'null' errors in Logstash.

Thanks @mpopov for creating the patch and walking me through the process! The instrument has been updated and I've confirmed we are no longer receiving validation errors for the user_edit_count_bucket field. Marking this to resolved.