Create mediawiki.user_change event stream
Open, HighPublic
Actions

Description

MediaWiki user change events are needed for Incremental MediaWiki History to support weekly delivery of core contributor metrics.

This Incremental MWH sheet tracks the data field requirements.

MediaWiki hooks available for use (as of 2026-04):

https://www.mediawiki.org/wiki/Manual:Hooks/LocalUserCreated
https://www.mediawiki.org/wiki/Manual:Hooks/UserGetRights
https://www.mediawiki.org/wiki/Manual:Hooks/UserGroupsChanged
https://www.mediawiki.org/wiki/Manual:Hooks/BlockIpComplete
https://www.mediawiki.org/wiki/Manual:Hooks/UnblockUserComplete
https://www.mediawiki.org/wiki/Manual:Hooks/RenameUserComplete
https://www.mediawiki.org/wiki/Manual:Hooks/UserSetEmail (maybe not needed?)
https://www.mediawiki.org/wiki/Manual:Hooks/UserSaveSettings (maybe not needed?)
https://www.mediawiki.org/wiki/Manual:Hooks/UserLoginComplete (probably not needed?)
https://www.mediawiki.org/wiki/Manual:Hooks/UserLogoutComplete (probably not needed?)
...?

Some prior work for producing this kind of event data is done in the ServerSideAccountCreation legacy eventlogging stream, emitted by Extension:Campaign. It looks like instead of using a convenient Hook, Extension:Campaign implements a SecondaryAuthenticationProvider that will be called by MediaWiki core when a user logs in. If the log in happens to also be an account creation, they construct an event and produce it.

There doesn't seem to be a relevant CentralAuth hook, so we'll have to figure out how global accounts relate.

Done is

new mediawiki/user/change(?) schema modeled on entity/user schema fragment
new mediawiki.user_change stream emitted from EventBus
new mediawiki.user_change has more signal than ServerSideAccountCreation does for Incremental MWH purposes

Details

Related Changes in Gerrit:

Subject	Repo	Branch	Lines +/-
stream: mediawiki.user_change	operations/mediawiki-config	master	+3 -11
Create new user_change event stream	mediawiki/extensions/EventBus	master	+1 K -6
EventStreamConfig - ingest mediawiki.user_change into the Data Lake	operations/mediawiki-config	master	+0 -11
EventStreamConfig - add mediawiki.user_change.dev0	operations/mediawiki-config	master	+31 -22

Customize query in gerrit

Related Changes in GitLab:

	Title	Reference	Author	Source Branch	Dest Branch
	Mediawiki user change - final location	repos/data-engineering/schemas-event-primary!60	javiermonton	feature/move-user-change-to-prod	master
	Add development/mediawiki_user_change	repos/data-engineering/schemas-event-primary!42	otto	user_change	master

Customize query in GitLab

Related Objects
Search...

Status	Assigned	Task
Open	None	T428273 Incremental MediaWiki History Phase II
Open	None	T291120 MediaWiki Event Carried State Transfer - Problem Statement
Open	Ottomata	T423935 Incremental MWH - MediaWiki event data source improvements
Open	Ottomata	T423952 Create mediawiki.user_change event stream
Open	None	T424685 Emit comprehensive mediawiki user block change information in an event stream
Open	Ottomata	T425360 EventBus - user entity schema should differentiate between explicit and implicit user groups
Resolved	Ottomata	T425801 CheckUser extension should use EventBus.UserEntitySerializer service instead of constructor
Open	Ottomata	T426198 Event schemas - mediawiki user entity should be wiki aware
Open	Ottomata	T424767 EventBus - consider schema versions when serializing entities
Open	None	T427817 MediaWiki DomainEvents - Create new User related DomainEvents
Open	None	T424887 mediawiki.page_change event - add access_method & platform details
Open	tchin	T425029 mediawiki.page_change.v1 - add revision.editor.first_edit_dt field
Open	Ottomata	T423583 mediawiki.page_change.v1 event - Add revision revert details
Open	None	T426998 mediawiki.page_change.v1 event - Add user first_registration_dt field
Open	None	T425996 Add log_id to mediawiki page_change and user_change events
Open	None	T427815 MediaWiki DomainEvents - Include LogEntry
Open	Ottomata	T428176 mediawiki.page_change.v1 events - add wikidata id for pages

Event Timeline

Ottomata created this task.Apr 20 2026, 8:36 PM

Ottomata added a parent task: T291120: MediaWiki Event Carried State Transfer - Problem Statement.

JAllemandou mentioned this in T423935: Incremental MWH - MediaWiki event data source improvements.Apr 21 2026, 7:09 AM

Ottomata updated the task description. (Show Details)Apr 22 2026, 4:41 PM

Ottomata updated the task description. (Show Details)Apr 22 2026, 5:12 PM

Met today with MW folks to learn more. Notes here.

Change #1277724 had a related patch set uploaded (by Ottomata; author: Ottomata):

[mediawiki/extensions/EventBus@master] WIP - create new user_change event stream

https://gerrit.wikimedia.org/r/1277724

gerritbot added a project: Patch-For-Review.Apr 27 2026, 7:57 PM

otto opened https://gitlab.wikimedia.org/repos/data-engineering/schemas-event-primary/-/merge_requests/42

Draft: Add development/mediawiki_user_change

Ottomata created subtask T424685: Emit comprehensive mediawiki user block change information in an event stream.Apr 28 2026, 5:19 PM

Ottomata mentioned this in T424685: Emit comprehensive mediawiki user block change information in an event stream.

Ottomata created subtask T424767: EventBus - consider schema versions when serializing entities.Apr 28 2026, 6:57 PM

Ottomata mentioned this in T424887: mediawiki.page_change event - add access_method & platform details.Apr 29 2026, 6:28 PM

Ottomata mentioned this in T424359: Architectural design agreement: Incremental MediaWiki History.Apr 29 2026, 6:42 PM

Ottomata mentioned this in T424767: EventBus - consider schema versions when serializing entities.Apr 30 2026, 6:44 PM

We might need to do {T348252} for this too.

Ottomata moved this task from Next Up to In progress on the Data-Engineering (Q4 FS25/26 April 1st - June 30st) board.May 4 2026, 3:04 PM

Ottomata created subtask T425360: EventBus - user entity schema should differentiate between explicit and implicit user groups.May 4 2026, 3:41 PM

Ottomata mentioned this in T425360: EventBus - user entity schema should differentiate between explicit and implicit user groups.

I just encountered an annoying issue with how we have been modeling user groups in event schemas. I filed T425360: EventBus - user entity schema should differentiate between explicit and implicit user groups. Fixing it may be a semi backwards incompatible change.

Since user groups are not needed for T418032, I'm going to not implement user group changes now. I have code that does it, so I'll separate that out.

For posterity, the user group change code I had can been seen by diffing patchset 8 to 7:
https://gerrit.wikimedia.org/r/c/mediawiki/extensions/EventBus/+/1277724/8..7

I worked out ^ as described in this comment: T425360#11887651

First draft of user_change ready! After review, we could merge and declare a .dev0 stream and try it!

Now to find a reviewer!

Ahoelzl added a project: DPE-MediaWiki-Incremental-History.May 4 2026, 11:34 PM

JAllemandou awarded a token.May 5 2026, 3:31 PM

I commented on the schema, but I'm not comfortable reviewing the mediawiki code :)

In T423952#11892934, @JAllemandou wrote:

I commented on the schema, but I'm not comfortable reviewing the mediawiki code :)

Same.

xcollazo mentioned this in T425573: mediawiki_history_incremental_v1: schema specification for stakeholder review.May 6 2026, 4:26 PM

Ottomata mentioned this in T425443: Mediawiki History Failure [2026-04].May 7 2026, 1:03 PM

Ottomata created subtask T425801: CheckUser extension should use EventBus.UserEntitySerializer service instead of constructor.May 8 2026, 5:09 PM

Ottomata mentioned this in T425801: CheckUser extension should use EventBus.UserEntitySerializer service instead of constructor.

Ottomata triaged this task as High priority.May 9 2026, 2:14 AM

Change #1285525 had a related patch set uploaded (by Ottomata; author: Ottomata):

[operations/mediawiki-config@master] EventStreamConfig - add mediawiki.user_change.dev0

https://gerrit.wikimedia.org/r/1285525

otto merged https://gitlab.wikimedia.org/repos/data-engineering/schemas-event-primary/-/merge_requests/42

Add development/mediawiki_user_change

Test wiki created on Patch demo by Ottomata using patch(es) linked to this task:
https://a199b51d8e.catalyst.wmcloud.org/w/

Change #1285525 merged by jenkins-bot:

[operations/mediawiki-config@master] EventStreamConfig - add mediawiki.user_change.dev0

https://gerrit.wikimedia.org/r/1285525

Mentioned in SAL (#wikimedia-operations) [2026-05-11T12:59:23Z] <otto@deploy1003> Started scap sync-world: Backport for [[gerrit:1285525|EventStreamConfig - add mediawiki.user_change.dev0 (T423952)]]

Mentioned in SAL (#wikimedia-operations) [2026-05-11T13:01:08Z] <otto@deploy1003> otto: Backport for [[gerrit:1285525|EventStreamConfig - add mediawiki.user_change.dev0 (T423952)]] synced to the testservers (see https://wikitech.wikimedia.org/wiki/Mwdebug). Changes can now be verified there.

Mentioned in SAL (#wikimedia-operations) [2026-05-11T13:07:28Z] <otto@deploy1003> Finished scap sync-world: Backport for [[gerrit:1285525|EventStreamConfig - add mediawiki.user_change.dev0 (T423952)]] (duration: 08m 05s)

Change #1277724 merged by jenkins-bot:

[mediawiki/extensions/EventBus@master] Create new user_change event stream

https://gerrit.wikimedia.org/r/1277724

ReleaseTaggerBot added a project: MW-1.47-notes (1.47.0-wmf.2; 2026-05-12).May 11 2026, 2:00 PM

Dreamy_Jazz closed subtask T425801: CheckUser extension should use EventBus.UserEntitySerializer service instead of constructor as Resolved.May 11 2026, 2:05 PM

Maintenance_bot removed a project: Patch-For-Review.May 11 2026, 2:31 PM

Ottomata moved this task from In progress to In Review on the Data-Engineering (Q4 FS25/26 April 1st - June 30st) board.May 11 2026, 3:01 PM

Ottomata moved this task from In Review to Ready to Deploy on the Data-Engineering (Q4 FS25/26 April 1st - June 30st) board.

A new consideration: T425986: Add log_id to wmf.mediawiki_history. For this new user_change stream we would also need a log_id field.

Change #1286434 had a related patch set uploaded (by Ottomata; author: Ottomata):

[operations/mediawiki-config@master] EventStreamConfig - ingest mediawiki.user_change into the Data Lake

https://gerrit.wikimedia.org/r/1286434

gerritbot added a project: Patch-For-Review.May 12 2026, 5:15 PM

Change #1286434 merged by jenkins-bot:

[operations/mediawiki-config@master] EventStreamConfig - ingest mediawiki.user_change into the Data Lake

https://gerrit.wikimedia.org/r/1286434

Mentioned in SAL (#wikimedia-operations) [2026-05-12T17:40:36Z] <otto@deploy1003> Started scap sync-world: Backport for [[gerrit:1286434|EventStreamConfig - ingest mediawiki.user_change into the Data Lake (T423952)]]

Mentioned in SAL (#wikimedia-operations) [2026-05-12T17:42:31Z] <otto@deploy1003> otto: Backport for [[gerrit:1286434|EventStreamConfig - ingest mediawiki.user_change into the Data Lake (T423952)]] synced to the testservers (see https://wikitech.wikimedia.org/wiki/Mwdebug). Changes can now be verified there.

Mentioned in SAL (#wikimedia-operations) [2026-05-12T17:56:44Z] <otto@deploy1003> Finished scap sync-world: Backport for [[gerrit:1286434|EventStreamConfig - ingest mediawiki.user_change into the Data Lake (T423952)]] (duration: 16m 08s)

Maintenance_bot removed a project: Patch-For-Review.May 12 2026, 6:30 PM

We have Hive event.mediawiki_user_change_dev0 records!

mszwarc mentioned this in T426185: TypeError: MediaWiki\Extension\EventBus\HookHandlers\MediaWiki\UserChangeHooks::calculateUserEffectiveGroups(): Argument #1 ($user) must be of type MediaWiki\User\User, MediaWiki\User\UserIdentityValue given, called in /srv/med.May 13 2026, 11:56 AM

Ottomata created subtask T426198: Event schemas - mediawiki user entity should be wiki aware.May 13 2026, 1:30 PM

In T423952#11915434, @Ottomata wrote:

We have Hive event.mediawiki_user_change_dev0 records!

Awesome! I'll hopefully start looking into these next week.

xcollazo mentioned this in T425729: Implement MERGE INTO writers for mediawiki_history_incremental_v1.Thu, May 21, 2:52 AM

Ottomata created subtask T427817: MediaWiki DomainEvents - Create new User related DomainEvents.Mon, Jun 1, 2:19 PM

Ottomata mentioned this in T427817: MediaWiki DomainEvents - Create new User related DomainEvents.

We have done multiple test runs ingesting data from event.mediawiki_user_change_dev0, and so far have found no data quality issues.

Can we move this dev stream to production?

CC @JMonton-WMF

javiermonton updated https://gitlab.wikimedia.org/repos/data-engineering/schemas-event-primary/-/merge_requests/60

Mediawiki user change - final location

Change #1299422 had a related patch set uploaded (by JavierMonton; author: JavierMonton):

[operations/mediawiki-config@master] stream: mediawiki.user_change

https://gerrit.wikimedia.org/r/1299422

gerritbot added a project: Patch-For-Review.Tue, Jun 9, 8:45 AM

I believe these are the patches needed to move it to "prod", and a restart of event gate I guess.

Thank you! Just discussed with Javier and Xabriel. We can do this, but I suggested to wait a while, perhaps until MWHInc v2.

We aren't yet sure how T424685: Emit comprehensive mediawiki user block change information in an event stream relates, and if it may change anything about the user/change data model if we decide to include it as part of the user_change stream.

Create mediawiki.user_change event streamOpen, HighPublicActions