Don't use async handler by Cali0707 · Pull Request #7415 · knative/eventing

Cali0707 · 2023-10-27T14:32:36Z

To try to improve the consistency of the system when using the IMC, this PR is testing switching to using the sync handler

Proposed Changes

Use the sync handler instead of the async handler

Pre-review Checklist

At least 80% unit test coverage
E2E tests for any new behavior
Docs PR for any user-facing impact
Spec PR for any new API feature
Conformance test for any change to the spec

Release Note

The InMemoryChannel now sends a 202 response only after successfully delivering the event to all subscribers

Docs

Signed-off-by: Calum Murray <cmurray@redhat.com>

Cali0707 · 2023-10-27T14:32:53Z

/cc @pierDipi
/hold
For testing a few times :)

codecov · 2023-10-27T14:41:15Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (0684dbe) 76.74% compared to head (0e701d3) 76.74%.
Report is 13 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #7415   +/-   ##
=======================================
  Coverage   76.74%   76.74%           
=======================================
  Files         253      253           
  Lines       13916    13916           
=======================================
  Hits        10680    10680           
  Misses       2702     2702           
  Partials      534      534

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Cali0707 · 2023-10-27T19:39:22Z

/retest

Signed-off-by: Calum Murray <cmurray@redhat.com>

Cali0707 · 2023-10-30T14:51:55Z

 			if dispatchResult.err != nil {
 				f.logger.Error("Fanout had an error", zap.Error(dispatchResult.err))
 				dispatchResultForFanout.err = dispatchResult.err
-				return dispatchResultForFanout


@pierDipi I think this early return may cause a race condition in the synchronous case, as it may lead to the context being cancelled (request completes) before all of the dead letter sink requests have been made if there is more than one error.

However, this means that we won't aggregate the errors and are only showing the final error. Should we add some kind of error aggregation here?

Error aggregation sounds good, we could track it into a separate issue

Signed-off-by: Calum Murray <cmurray@redhat.com>

Cali0707 · 2023-10-30T20:29:45Z

 	env.Prerequisite(ctx, t, broker.GoesReady("default", b.WithEnvConfig()...))
-	env.TestSet(ctx, t, broker.ControlPlaneConformance("default", b.WithEnvConfig()...))
 	env.TestSet(ctx, t, broker.DataPlaneConformance("default"))
+	env.TestSet(ctx, t, broker.ControlPlaneConformance("default", b.WithEnvConfig()...))


@pierDipi changing the order of these tests caused the DataPlaneConformance tests to pass 😕

Cali0707 · 2023-10-30T21:17:33Z

/retest

Cali0707 · 2023-10-31T12:59:07Z

/test upgrade-tests
(this is what we are trying to improve)

Cali0707 · 2023-11-01T18:49:23Z

/test upgrade-tests
(retesting again to see if the flakes are gone)

pierDipi · 2023-11-14T10:53:12Z

3/3 green on the last commit (retesting again to see if the flakes are gone)
/test upgrade-tests

Cali0707 · 2023-11-14T13:24:52Z

Still green, going to try again for the flakes:
/test upgrade-tests

Cali0707 · 2023-11-14T16:23:25Z

/test upgrade-tests

Cali0707 · 2023-11-14T16:57:33Z

/unhold

pierDipi

Great!

/lgtm
/approve

knative-prow · 2023-11-14T17:00:26Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: Cali0707, pierDipi

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

~~OWNERS~~ [Cali0707,pierDipi]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

* Don't use async handler Signed-off-by: Calum Murray <cmurray@redhat.com> * Fix Kn-Namespace header, potential race condition for sync receiver Signed-off-by: Calum Murray <cmurray@redhat.com> * try refactoring test Signed-off-by: Calum Murray <cmurray@redhat.com> * take 2 on test fix Signed-off-by: Calum Murray <cmurray@redhat.com> --------- Signed-off-by: Calum Murray <cmurray@redhat.com>

* Don't use async handler * Fix Kn-Namespace header, potential race condition for sync receiver * try refactoring test * take 2 on test fix --------- Signed-off-by: Calum Murray <cmurray@redhat.com>

Don't use async handler

8dd3e6b

Signed-off-by: Calum Murray <cmurray@redhat.com>

knative-prow Bot requested review from aliok and odacremolbap October 27, 2023 14:32

knative-prow Bot added approved Indicates a PR has been approved by an approver from all required OWNERS files. size/XS Denotes a PR that changes 0-9 lines, ignoring generated files. labels Oct 27, 2023

knative-prow Bot requested a review from pierDipi October 27, 2023 14:32

knative-prow Bot added the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label Oct 27, 2023

Fix Kn-Namespace header, potential race condition for sync receiver

ff97c1d

Signed-off-by: Calum Murray <cmurray@redhat.com>

Cali0707 commented Oct 30, 2023

View reviewed changes

knative-prow Bot added size/S Denotes a PR that changes 10-29 lines, ignoring generated files. area/test-and-release Test infrastructure, tests or release and removed size/XS Denotes a PR that changes 0-9 lines, ignoring generated files. labels Oct 30, 2023

Cali0707 force-pushed the use-sync-handler-imc branch from d02844b to ff97c1d Compare October 30, 2023 16:13

try refactoring test

54078cd

Signed-off-by: Calum Murray <cmurray@redhat.com>

Cali0707 force-pushed the use-sync-handler-imc branch from cf33d77 to 54078cd Compare October 30, 2023 18:50

take 2 on test fix

0e701d3

Signed-off-by: Calum Murray <cmurray@redhat.com>

knative-prow Bot added size/XS Denotes a PR that changes 0-9 lines, ignoring generated files. and removed size/M Denotes a PR that changes 30-99 lines, ignoring generated files. labels Oct 30, 2023

Cali0707 commented Oct 30, 2023

View reviewed changes

knative-prow Bot removed the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label Nov 14, 2023

pierDipi approved these changes Nov 14, 2023

View reviewed changes

knative-prow Bot assigned pierDipi Nov 14, 2023

knative-prow Bot added the lgtm Indicates that a PR is ready to be merged. label Nov 14, 2023

knative-prow Bot merged commit d964da3 into knative:main Nov 14, 2023

Cali0707 mentioned this pull request Nov 15, 2023

Investigate more graceful shutdown of the InMemoryChannel #7407

Closed

This was referenced Dec 7, 2023

Cleanup created resources in channel conformance tests between features #7500

Merged

[Automated] Update eventing-istio-eventing nightly knative-extensions/eventing-istio#158

Merged

Conversation

Cali0707 commented Oct 27, 2023

Proposed Changes

Pre-review Checklist

Uh oh!

Cali0707 commented Oct 27, 2023

Uh oh!

codecov Bot commented Oct 27, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Cali0707 commented Oct 27, 2023

Uh oh!

Cali0707 Oct 30, 2023

Choose a reason for hiding this comment

Uh oh!

pierDipi Nov 14, 2023

Choose a reason for hiding this comment

Uh oh!

Cali0707 Oct 30, 2023

Choose a reason for hiding this comment

Uh oh!

Cali0707 commented Oct 30, 2023

Uh oh!

Cali0707 commented Oct 31, 2023

Uh oh!

Cali0707 commented Nov 1, 2023

Uh oh!

pierDipi commented Nov 14, 2023

Uh oh!

Cali0707 commented Nov 14, 2023

Uh oh!

Cali0707 commented Nov 14, 2023

Uh oh!

Cali0707 commented Nov 14, 2023

Uh oh!

pierDipi left a comment

Choose a reason for hiding this comment

Uh oh!

knative-prow Bot commented Nov 14, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

codecov Bot commented Oct 27, 2023 •

edited

Loading