[Feature] should fix the version of gprc

[ShuHu2015]

Search before asking

• I had searched in the issues and found no similar feature requirement.

Description

I tried python apache-skywalking version 1.0.1 to 1.2.0, all of them raised the grpc error with grpc version 1.83.0 which updated last week. here is the errors:
E0728 09:36:09.293938 496 ev_poll_posix.cc:850] pollset_kick_ext: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.293967 496 completion_queue.cc:800] Kick failed: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.434072 111 ev_poll_posix.cc:850] pollset_kick_ext: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.434106 111 completion_queue.cc:800] Kick failed: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.434641 424 ev_poll_posix.cc:850] pollset_kick_ext: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.434682 424 completion_queue.cc:800] Kick failed: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.469885 581 ev_poll_posix.cc:850] pollset_kick_ext: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.469917 581 completion_queue.cc:800] Kick failed: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.504783 228 ev_poll_posix.cc:850] pollset_kick_ext: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.504817 228 completion_queue.cc:800] Kick failed: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.541328 364 ev_poll_posix.cc:850] pollset_kick_ext: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.541370 364 completion_queue.cc:800] Kick failed: UNKNOWN:Kick Failure (eventfd_write: Bad file descriptor (9))
E0728 09:36:09.565696

Use case

please fixed the grpc version or compatible grpc version 1.83.0

Related issues

No response

Are you willing to submit a pull request to implement this on your own?

• Yes I am willing to submit a pull request on my own!

Code of Conduct

• I agree to follow this project's Code of Conduct

[wu-sheng]

Which oap version did you use? Whether it is latest 10.x or 11 dev?

[wu-sheng]

Our e2e at both repo(python and main), I am not sure how to reproduce your error.

[ShuHu2015]

Our e2e at both repo(python and main), I am not sure how to reproduce your error.

pip install grpcio with 1.83.0 version.
grpcio==1.83.0
grpcio-tools==1.83.0
And try this: sw-python run -p gunicorn --bind :5002 -w $workers -k gthread --thread=$thread my_app:app

[wu-sheng]

Which oap version with which storage?

[ShuHu2015]

Which oap version with which storage?

9.7.0

[wu-sheng]

We finished the root-cause analysis and the fix. Summary of what is actually happening:

This is not related to the OAP version. The Kick Failure (eventfd_write: Bad file descriptor) errors come entirely from the gRPC client's polling engine at fork() time — we reproduced the identical output against a mock collector with no OAP involved at all. (Upgrading from 9.7.0 is still a good idea, but for other reasons.)

It is a grpc-python fork regression, and it is not specific to 1.83.0. gRPC Python has been migrating to the EventEngine poller since grpcio 1.80.0; in our tests grpcio 1.82.1 produces the same error spam in the same Gunicorn scenario. What changed in 1.83.0 (grpc/grpc#42828) is that the last opt-out was removed, so the legacy fork recipe (GRPC_ENABLE_FORK_SUPPORT + GRPC_POLL_STRATEGY=poll) that the Python agent applied in prefork mode — and that gRPC's own fork_support.md still documents — no longer works. See upstream grpc/grpc#43055 and grpc/grpc#43062, both open without a fix.

Good news: in all of our reproductions, tracing kept working despite the errors — workers served traffic and every trace segment was delivered. The spam is stale-file-descriptor noise. However, we also found a worse latent failure mode on some grpcio versions: a forked worker can occasionally deadlock silently inside gRPC's own at-fork handlers (the worker never boots and Gunicorn does not notice), so this needed a structural fix, not a version pin.

The fix is in apache/skywalking-python#409. The root of every failure mode is a live gRPC channel in the Gunicorn master at fork() time, so the agent no longer creates one: with sw-python run -p gunicorn, the master now only installs instrumentation, and the full agent (including the gRPC channel) starts in each forked worker — the same model the uWSGI path has always used. This was verified clean on grpcio 1.80.0 / 1.82.1 / 1.83.0 with repeated fork cycles. The fixed agent requires grpcio >= 1.83 (the generated protobuf stubs already require it at import time).

Workarounds for released agents (<= 1.2.0) until the next release:

• pin grpcio<1.80 (the last series fully clean of the fork regression), or
• switch the reporter to SW_AGENT_PROTOCOL=http (or kafka), which avoids gRPC entirely.