send headers when failure is null #335
Conversation
|
|
|
@jamesward @lowasser hi, do you have any discussion about this? |
|
It's not entirely clear to me under what circumstances this would happen. Can you add a test that fails before and passes now? |
|
@jamesward @lowasser hi, i add test example. Sorry, the test sample code is not elegant. |
|
Are retries the only setting in which this issue happens? I'll admit I guessed that this would just be a particular arrangement of failures in a normal RPC, but I can believe gRPC handles retries in a special way that can't be otherwise replicated. It also looks like you added AbstractCoroutinesInteropTest to main, when it should just live in test. |
|
@lowasser I haven't tried everything, there seems to be no problem with other settings. thank you. 😄 |
|
I believe that this PR would solve it, but it'd definitely help if we could simplify the tests so it's easier to follow how the bug occurs. You'd also need to move the tests into the correct directories, at least, before we can merge this. |
|
@lowasser hi, i add test case |
| @@ -0,0 +1,265 @@ | |||
| /* | |||
There was a problem hiding this comment.
This file shouldn't be getting added at all.
There was a problem hiding this comment.
remove this file.
how about another test case. (in ServerCallTest$coroutinesServerRetry)
|
@lowasser Hi, Have you seen my edits? |
|
@lowasser Hi, do you by any chance remember this? |
Hello.
I found an issue that didn't work when I tried retry with coroutines.
I think the problem with it is due to sendHeaders.
So, I added a condition to send only when failure is null.
If this is not the reason, you may close this PR.
thank you.
Related issues: #334