Skip to content

Conversation

@anoopcs9
Copy link
Collaborator

We previously switched from official CentOS vagrant boxes as they were not regularly updated. But the situation has improved a lot since the last year where boxes are more frequently updated. On top of all that vagrant cloud platform(HCP-Hashicorp Cloud Platform) has recently implemented rate limits for downloading boxes. Given all these facts it is better to get back to official CentOS vagrant boxes. These are externally hosted on CentOS cloud(and properly redirected from HCP) allowing us to bypass the throttling seen with HCP.

Copy link
Collaborator

@phlogistonjohn phlogistonjohn left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

seems reasonable to me.

@anoopcs9
Copy link
Collaborator Author

We previously switched from official CentOS vagrant boxes as they were not regularly updated. But the situation has improved a lot since the last year where boxes are more frequently updated. On top of all that vagrant cloud platform(HCP-Hashicorp Cloud Platform) has recently implemented rate limits for downloading boxes. Given all these facts it is better to get back to official CentOS vagrant boxes. These are externally hosted on CentOS cloud(and properly redirected from HCP) allowing us to bypass the throttling seen with HCP.

Hashicorp confirms that it was misconfiguration. Anyway I'll try to fix the CI failures with the switch to CentOS official boxes.

@anoopcs9 anoopcs9 force-pushed the use-vagrant-box-centos-cloud branch 2 times, most recently from eab818d to b96a611 Compare November 26, 2025 11:50
@anoopcs9
Copy link
Collaborator Author

/retest all

We don’t want any surprises while resolving dependencies for any Samba
packages that may be present by default.

Signed-off-by: Anoop C S <[email protected]>
We previously switched[1] from official CentOS vagrant boxes as they
were not regularly updated. But the situation has improved a lot since
the last year where boxes are more frequently updated. On top of all
that vagrant cloud platform(HCP-Hashicorp Cloud Platform) has recently
implemented rate limits for downloading boxes[2]. Given all these facts
it is better to get back to official CentOS vagrant boxes. These are
externally hosted on CentOS cloud(and properly redirected from HCP)
allowing us to bypass the throttling seen with HCP.

[1] samba-in-kubernetes@d1a8534
[2] hashicorp/vagrant#13590

Signed-off-by: Anoop C S <[email protected]>
@anoopcs9 anoopcs9 force-pushed the use-vagrant-box-centos-cloud branch from b96a611 to b2f6a45 Compare November 26, 2025 18:27
@xhernandez
Copy link
Collaborator

It looks like the new images already have a newer kernel than the one we want to install and it doesn't get replaced. We should force the installation of the kernel, even if it's a downgrade. I'll try to look at this, maybe tomorrow.

@anoopcs9
Copy link
Collaborator Author

It looks like the new images already have a newer kernel than the one we want to install and it doesn't get replaced.

Yes, I had noticed it.

We should force the installation of the kernel, even if it's a downgrade. I'll try to look at this, maybe tomorrow.

Ok, feel free to update the branch with an additional commit.

@anoopcs9 anoopcs9 marked this pull request as ready for review November 28, 2025 10:25
@anoopcs9
Copy link
Collaborator Author

We should force the installation of the kernel, even if it's a downgrade. I'll try to look at this, maybe tomorrow.

Ok, feel free to update the branch with an additional commit.

Thanks for the minimal change 🙂.

@anoopcs9 anoopcs9 requested a review from xhernandez November 28, 2025 10:26
@xhernandez xhernandez merged commit ec97ae9 into samba-in-kubernetes:main Nov 28, 2025
7 checks passed
@anoopcs9 anoopcs9 deleted the use-vagrant-box-centos-cloud branch November 28, 2025 10:31
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.

3 participants