Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

set http2 to default false #161

Closed
wants to merge 1 commit into from
Closed

Conversation

akdienes
Copy link

@akdienes akdienes commented May 2, 2023

PrefectHQ/prefect#7442
encode/httpx#2112

considering that the above two issues mean that using http2: true leads consistently to irrecoverable worker failures, I think the default should be false

PrefectHQ/prefect#7442
encode/httpx#2112

considering that the above two issues mean that using `http2: false` leads consistently to irrecoverable worker failures, I think the default should be false
@akdienes akdienes requested review from gabcoyne, zanieb and a team as code owners May 2, 2023 16:45
@jawnsy
Copy link
Contributor

jawnsy commented May 2, 2023

Hey there! Thanks for your pull request.

There is a performance cost associated with this change, and the irrecoverable worker failures should be resolved with PrefectHQ/prefect#9267

Are you running the latest version of this chart?

@akdienes
Copy link
Author

akdienes commented May 2, 2023

Hi @jawnsy , thanks for the quick response!

I was about 2 weeks behind latest, which looks like right around when that PR was merged. I'll update that and let you know if the worker failures persist or not

@akdienes
Copy link
Author

akdienes commented May 2, 2023

just wanted to update that I attempted to update the chart, but it causes the worker to fail in kube with
prefect.exceptions.PrefectHTTPStatusError: Client error '404 Not Found' for url 'https://prefect.manifoldtrading.net/api/deployments/89ee0166-b731-4967-8d0b-19d7d3a6bb2e'

Indeed this deployment does not exist. I'm not sure where it's getting that from or why it's trying to query that. moreover, the prefect UI shows the worker as healthy despite crashing repeatedly in kube

anyway, I've reverted back to the chart from mar 10 and the above error as gone away. just setting http2: false locally for the time being

@jawnsy
Copy link
Contributor

jawnsy commented May 30, 2023

This was an issue with an upstream dependency, which should have been resolved in encode/httpcore 0.17.1. I'm closing this out as the default behavior (to use HTTP/2) should now work reliably. Please let us know if you see otherwise after upgrading to the latest version.

@jawnsy jawnsy closed this May 30, 2023
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.

2 participants