Critical Bug Fix | Nested generated Distributed Event is not being published

[alyssonrpg]

Description

This pull request fixes a bug related to the publication of distributed events when they are generated in a nested manner during the completion of a UnitOfWork.

Scenario in Which the Bug Occurs

1. Initiating a UnitOfWork: A new UnitOfWork is started.

2. Appending Events: Both a Local Event and a Distributed Event are appended to the UnitOfWork.

3. Completing the UnitOfWork: The CompleteAsync method is called on the UnitOfWork.

4. Handling the Local Event: The ILocalEventHandler for the local event is invoked.

5. Nested Generation of a Distributed Event: During the execution of the ILocalEventHandler, a new Distributed Event is generated in a nested manner.

Identified Issue

• The Distributed Event generated in step 5 is not being published to the IUnitOfWorkEventPublisher.
• This occurs because the loop that publishes local and distributed events in UnitOfWork.CompleteAsync() is not correctly handling this nestedly generated event.
• As a result, there is a critical miss in event publication, leading to potential inconsistencies in the system and affecting functionalities that rely on the proper propagation of distributed events.

Checklist

• I fully tested it as developer / designer and created unit / integration tests
• I documented it (or no need to document or I will create a separate documentation issue)

How to Test It?

I have included a unit test in this pull request: Volo.Abp.Uow.UnitOfWork_Event_Publisher_Tests that tests this scenario. To observe the bug, simply run this test with the previous version of UnitOfWork.cs to see the issue occur.

[CLAassistant]

All committers have signed the CLA.

[maliming]

Thanks @alyssonrpg

I have moved the code to the rel-9.0 branch. The dev branch is targeted to 9.2.

#22056

[alyssonrpg]

Thanks @alyssonrpg

I have moved the code to the rel-9.0 branch. The dev branch is targeted to 9.2.

#22056

@maliming Nice! However, I'd like to suggest using the unit test I've written in this PR as well. The test in PR #22056 would fail to identify this bug.

The bug happens when a local event handler, invoked within that loop, attempts to publish a new distributed event.

[maliming]

hi @alyssonrpg
I update the test code. It is the same as yours.

https://github.com/abpframework/abp/pull/22056/files#diff-fc97f9f91f2d3caf7c4ea9d078dadd373d3953ccfddd26b50793c5180ec0ef80R182-R226

[alyssonrpg]

hi @alyssonrpg I update the test code. It is the same as yours.

https://github.com/abpframework/abp/pull/22056/files#diff-fc97f9f91f2d3caf7c4ea9d078dadd373d3953ccfddd26b50793c5180ec0ef80R182-R226

@maliming sorry, my bad. I am reading the code on a phone screen and I missed a detail. The test is alright