-
-
Notifications
You must be signed in to change notification settings - Fork 1.8k
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
Add CI compose version 1.29.2 / 2.2.3 #1251
Add CI compose version 1.29.2 / 2.2.3 #1251
Conversation
Revert? 👀 |
Trying to see if the problem happens again with v2.0.1 or what? Sorry for the noise. |
Ah, gotcha. I figured it was #1171. :-/ |
I'm running self-hosted (21.11.0) with |
OK, So I went and checked this PR a couple of times during the past few days and couldn't figure out what is wrong with 2.2.3, last solution I have is to check every release between 2.0.1 and 2.2.3 to see what breaks where :) and then figure out why CI Passed:
So the regression starts with 2.2.3 (https://github.com/docker/compose/releases/tag/v2.2.3) itself. |
|
@aminvakil could this be some sort of Unix/Windows line ending issue? |
@aminvakil, actually, I bet self-hosted/install/relay-credentials.sh Lines 16 to 20 in 13d14e8
Producing an invalid YAML, causing relay to fail. That means getsentry/relay#1169 is invalid and we should address the issue here or with |
@BYK are there any issues with not using I think the bug described in the comment above might also purely exist because you're redirecting stdout to a file instead of letting Relay create the file, because that creates an empty file before the relay process starts. |
@untitaker using a volume is fine here but Relay does not like that scenario: getsentry/relay#564 |
I removed putting output to file to see what is the output in 6366592. I guess it's related to docker/compose#9035 which fixed another bug reported by @chadwhitacre :) |
That linked PR only changes relay's behavior when using the |
Closing and reopening to see if kafka thing is temporary or not. |
@aminvakil let's try what @untitaker suggest and remove the |
Oh wow I've been missing the party. 🍿 |
@BYK @chadwhitacre Sorry for answering late. I wanted to give myself more time to think about this.
I propose keeping this PR as it is, but create another PR which changes relay credentials generation as @untitaker suggests, and take a look at this PR after that has been changed, what do you think?
Sure, why not? I say add them later after merging the relay credential PR. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If -T
fixes things, I actually like this approach more. Should we bake that into the $dcr
shortcut?
Also, we either need to add the old versions to the matrix too or update the required checks in branch protection rules @chadwhitacre
I'm not sure how it would act on older docker-compose versions or different parts which |
Then please explain why that is needed on the |
b36708d
to
f34f46f
Compare
637c429
to
a1c91ae
Compare
Last failed run was because of docker/compose#6704 |
a1c91ae
to
32e1975
Compare
Interesting, docker/compose#9035, which introduced the need for |
32e1975
to
1b5eb5c
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I dig it! 👍
👏 👏 @aminvakil as well as @BYK for wrestling this to the ground. It was supposed to be easy! 🙈 |
@BYK Thank you for helping me through this! |
This reverts commit 3abbb3e.
If it's causing issues (#1133 (comment)) why should we stick to it?Also while we're here let's bump docker-compose 1.x version to latest release.So this PR got changed in the way and at last min/max 1.x and 2.x versions of docker-compose added to CI as suggested by @chadwhitacre in #1251 (comment)
Also because of recent changes in 2.2.3 (docker/compose#9035)
-T
had to be added to some scripts using$dcr
where output of it was getting used.