On Jule 30, 2024 many folks faced inaccessible repositories in the CentOS Vault.

We faced the same situation when CentOS Linux 8 went EOL. Back then @bstinson came up with a solution that involved using Cloudfront caching (https://pagure.io/centos-infra/issue/682).

It seems that the Cloudfront caching is not a sufficient solution anymore.

This kind of an outage has a business impact on Red Hat as the company supports conversions from CentOS Linux 7 to RHEL 7 and:
- currently we tell customers to use the CentOS Vault when performing the conversion
- we have an KCS article guiding customers to switch to the Vault
- we provide official Ansible roles for converting to RHEL that switch the CentOS repos to the Vault
- the RHEL Conversions team uses the CentOS Vault in automated integration tests

There's a workaround - to point to a third party repo in all the above places. But Red Hat customers who are already in preparations to do the conversion have most likely set up their automation to switch the CentOS repos to the Vault based on our guidance and those will start seeing errors.

If possible, please make the infrastructure more robust so it can handle higher loads.