Skip to content

Conversation

@yim-lee
Copy link
Member

@yim-lee yim-lee commented Jun 8, 2022

Resolves #947

@yim-lee
Copy link
Member Author

yim-lee commented Jun 8, 2022

Tests pass but there are a bunch of these in test logs:

23:56:13 2022-06-08T06:56:13+0000 error /system/receptionist : actor/id=/system/receptionist actor/path=/system/receptionist Error: timedOut(DistributedActors.TimeoutError(message: "Remote call timed out", timeout: TimeAmount(5s, nanoseconds: 5000000000)))
23:56:13 ERROR: Cannot schedule tasks on an EventLoop that has already shut down. This will be upgraded to a forced crash in future SwiftNIO versions.

They are not related.

One can reproduce the EL-already-shut-down errors by running ClusterSystemTests alone (even just the ClusterSystemTests.test_remoteCall subset would do).

The receptionist error might have been caused by the changes in InvocationBehavior.

@ktoso
Copy link
Member

ktoso commented Jun 8, 2022

I dug deep into this and it will need some timer scheduling changes -- too large to do in this PR.

We figured it out with @Lukasa and I'll document a plan in a new ticket.

@ktoso ktoso merged commit 9bfc901 into apple:main Jun 8, 2022
@yim-lee yim-lee deleted the iss947 branch June 8, 2022 17:37
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Implement remoteCall without ask actors, only send a "call ID" and store it

2 participants