feat: allows hook in the subscriptions #1309
Draft
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…resolve at the initialization of each client subscription
…duce duplicated code
…ce to fix when a subscription is completed while a hook is still sending messages)
…bscriptiononstarthandler
…tarthandler' into ale/eng-7600-add-subscriptiononstarthandler
…urce subscription
…iption to the same trigger
… has been migrated to router
…trigger in GraphQL
…ng-7601-add-streambatcheventhook-and-streampublisheventhook
|
Important Review skippedDraft detected. Please check the settings in the CodeRabbit UI or the You can disable this status message by setting the Comment |
<!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Tests** * Enhanced subscription-related test coverage for the GraphQL data source, aligning test utilities with current interfaces to improve reliability and maintainability. * Validated subscribe/unsubscribe and update flows under various scenarios to reduce brittleness and increase confidence in behavior. * Performed minor cleanup in test code to improve readability without altering functionality. <!-- end of auto-generated comment: release notes by coderabbit.ai --> ## Checklist - [ ] I have discussed my proposed changes in an issue and have received approval to proceed. - [ ] I have followed the coding standards of the project. - [ ] Tests or benchmarks have been added or updated. Implements missing methods to satisfy the `resolve.SubscriptionUpdater` interface, so tests can build. Since these methods are not used by the tests, they are empty. There tests in place already in resolve_test.go, which test wether the updater can manage subscriptions individually, so I did not add any.
@coderabbitai summary ## Checklist - [ ] I have discussed my proposed changes in an issue and have received approval to proceed. - [ ] I have followed the coding standards of the project. - [ ] Tests or benchmarks have been added or updated. I adjusted two tests which are failing during Githubs CI run: ``` --- FAIL: TestResolver_ResolveGraphQLSubscription (31.69s) --- FAIL: TestResolver_ResolveGraphQLSubscription/SubscriptionOnStart_ctx_updater_only_updates_the_right_subscription (0.01s) resolve_test.go:5604: Error Trace: /home/runner/work/graphql-go-tools/graphql-go-tools/v2/pkg/engine/resolve/resolve_test.go:5604 Error: Should be true Test: TestResolver_ResolveGraphQLSubscription/SubscriptionOnStart_ctx_updater_only_updates_the_right_subscription --- FAIL: TestResolver_ResolveGraphQLSubscription/SubscriptionOnStart_ctx_updater_on_multiple_subscriptions_with_same_trigger_works (0.01s) resolve_test.go:5674: Error Trace: /home/runner/work/graphql-go-tools/graphql-go-tools/v2/pkg/engine/resolve/resolve_test.go:5674 Error: should not be here Test: TestResolver_ResolveGraphQLSubscription/SubscriptionOnStart_ctx_updater_on_multiple_subscriptions_with_same_trigger_works FAIL ``` These tests work locally, at least on my machine. I assumed it only manifests itself on slower CPU´s like it's the case on Github workers. After testing around with some time.Sleeps() here and there, I found that if messages in tests are not emitted fast enough, then heartbeat messages might find their way into the recorders message list. As these tests have nothing to do with heartbeat testing I disabled them here. I am not 100% sure this is the problem, but I am fairly certain and I think it makes sense anyway.
@coderabbitai summary ## Checklist - [x] I have discussed my proposed changes in an issue and have received approval to proceed. - [x] I have followed the coding standards of the project. - [x] Tests or benchmarks have been added or updated. # Context Some notes first: - Its only happening for tests introduced on the cosmo streams topic branch - It seems to be a race condition in tests rather than actual engine code I spotted two tests failing on Github Actions due to race conditions. They work locally and are CPU timings related. Those two tests are - test 1 `SubscriptionOnStart ctx updater only updates the right subscription` - test 2 `SubscriptionOnStart ctx updater on multiple subscriptions with same trigger works` ### test 1: There is a race condition going on. Here is the output of the test on Github runners with engine logs enabled. ``` resolver:trigger:subscription:add:17241709254077376921:1 resolver:create:trigger:17241709254077376921 resolver:trigger:start:17241709254077376921 resolver:subscription_updater:update:17241709254077376921 resolver:trigger:initialized:17241709254077376921 resolver:subscription_updater:update:17241709254077376921 resolver:trigger:subscription:update:17241709254077376921:1,1 resolver:trigger:update:17241709254077376921 resolver:trigger:subscription:add:17241709254077376921:2 resolver:trigger:subscription:added:17241709254077376921:2 resolver:trigger:subscription:update:1 resolver:trigger:subscription:flushed:1 resolver:trigger:subscription:update:1 resolver:trigger:subscription:flushed:1 resolver:trigger:started:17241709254077376921 resolver:subscription_updater:complete:17241709254077376921 resolver:subscription_updater:complete:sent_event:17241709254077376921 resolver:trigger:complete:17241709254077376921 resolver:trigger:complete:17241709254077376921 resolver:trigger:subscription:closed:17241709254077376921:1 resolver:trigger:subscription:closed:17241709254077376921:2 recorder 1 messages: [{"data":{"counter":1000}} {"data":{"counter":0}}] recorder 2 messages: [] ``` As you can see recorder 2 misses its one expected message. The reason is that we update the trigger with the counter=0 message (line 8) before the second subscriber is added (line 9). So it misses the message. This happens because in the test we don't wait for the subscriber to finish registration on the trigger before sending the counter=0 message. Now we actually wait for that. ### test 2: Kind of the same error. Here is the engine debug output from a failing Github Actions run: ``` resolver:trigger:subscription:add:15889878720417707388:1 resolver:create:trigger:15889878720417707388 resolver:trigger:start:15889878720417707388 resolver:subscription_updater:update:15889878720417707388 resolver:trigger:initialized:15889878720417707388 resolver:subscription_updater:update:15889878720417707388 resolver:trigger:subscription:update:15889878720417707388:1,1 resolver:trigger:update:15889878720417707388 resolver:trigger:subscription:add:15889878720417707388:2 resolver:trigger:subscription:added:15889878720417707388:2 resolver:subscription_updater:update:15889878720417707388 resolver:trigger:subscription:update:15889878720417707388:1,2 resolver:trigger:subscription:update:2 resolver:trigger:started:15889878720417707388 resolver:trigger:subscription:update:1 resolver:trigger:subscription:flushed:2 resolver:trigger:subscription:flushed:1 resolver:trigger:subscription:update:1 resolver:trigger:subscription:flushed:1 resolver:subscription_updater:complete:15889878720417707388 resolver:subscription_updater:complete:sent_event:15889878720417707388 resolver:trigger:complete:15889878720417707388 resolver:trigger:complete:15889878720417707388 resolver:trigger:subscription:closed:15889878720417707388:1 resolver:trigger:subscription:closed:15889878720417707388:2 recorder 1 messages: [{"data":{"counter":1000}} {"data":{"counter":0}}] recorder 2 messages: [{"data":{"counter":1000}}] ``` As you can see recorder 2 misses the counter=0 message. Both are expected to have the same messages in the same order. Both recorders have the counter=1000 message, which is delivered via subscription-on-start hook but recorder 2 misses the counter=0 message, delivered via fake stream. The count=0 message is delivered (line 8) before recorder 2 is subscribed (line 9). This happens because in this test, like in the other, we don't wait for the recorders to finish subscribing to the trigger, and sending off the counter=0 messages via fake stream early. Its fixed by waiting for a complete subscription.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR introduces hooks inside the subscription lifecycle. We also decided to remove old pubsub implementation that is already deprecated and the router is not using anymore.
@coderabbitai summary
Checklist