Add OpenAI extension #8227
Add OpenAI extension #8227
Conversation
… check-extension-pages mojo
| .statusCode(204); | ||
|
|
||
| // Assert the streamed results | ||
| IntStream.rangeClosed(1, 10) |
There was a problem hiding this comment.
Do we rely on message ordering for expectation ? If yes, not sure the route would preserve order, not sure of the opposite though.
There was a problem hiding this comment.
We do rely on ordering for this test...
Unless the LLM gets the order of numbers wrong (unlikely), won't the streamed chunks come in the expected 1-10 order?
Or do you think the messages arriving at the seda endpoint might be out of order (I've not seen it happen so far in local testing)?
There was a problem hiding this comment.
I was more fearing a kind of race condition with wireTap thread pool but nothing demonstrated. If we could not simply switch to non order expectations now, we are not taking a big risk anyway. Even if flaky, it would not happen a lot 🤞 .
There was a problem hiding this comment.
Ok - thanks for the feedback. Lets see how it goes. If it does prove to be flaky on CI, I will rework it.
There was a problem hiding this comment.
@aldettinger I think you are right about this!
Seems the latest Windows failure is due to unexpected ordering.
I'll try to rework the test to make it more reliable.
There was a problem hiding this comment.
Ok, that said the probability should have been quite low. Anyway, no order in the route is not a big deal, just need to relax assertion.
aldettinger
left a comment
aldettinger
left a comment
There was a problem hiding this comment.
Just a few neat questions. The comments and doc are really helpful for review 👏
|
I need to fix things up for Windows where I added test file deletion: |
jamesnetherton
force-pushed
the
add-extension-openai
branch
from
3e18d12 to
dd05e3a
Compare
jamesnetherton
force-pushed
the
add-extension-openai
branch
from
dd05e3a to
f49327d
Compare
3 participants
The
openaicomponent is not in the 4.17.0 catalog. This scenario happens occasionally, so I added a small 'enhancement' to thecheck-extension-pagesmojo. So that it can continue to generate the expected doc resources. I intend to immediately remove the additional plugin config once this change lands on thecamel-mainbranch.