Broken CDN worker for repo.anaconda.com

[kalebo]

Checklist

• I added a descriptive title
• I searched open reports and couldn't find a duplicate

What happened?

One of the cloudflare CDN workers for repo.anaconda.com seems to be broken for the Salt Lake City region. It returns 500 error codes for GET requests both within conda, as well via curl or a browser. It has been doing this for about 4 hours at this point.

Specifically, the error output is:

CondaHTTPError: HTTP 500 INTERNAL SERVER ERROR for url <https://repo.anaconda.com/pkgs/main/linux-64/repodata.json>
Elapsed: 00:06.508551
CF-RAY: 8ece52c17e992c69-SLC

Additional Context

No response

[chenghlee]

Checked with Anaconda's infrastructure team: we identified a spike in HTTP 500 errors affecting ~1% of CDN traffic for repo.anaconda.com, starting around 2024-12-04 1600 UTC. The problem appears to be isolated to certain regions (so far: Detroit, Malaysia, and possibly Utah), and the issue seems to be on our CDN provider's side (CloudFlare). We've opened a ticket with them and will update the community as we get more information.

[jezdez]

We've opened an incident and published it on Anaconda's status page: https://anaconda.statuspage.io/incidents/qfv2n662hyym

[jezdez]

It seems like the issue was found by the infrastructure provider and a fix is rolling out at the moment

[jezdez]

Closing as completed following confirmation by Anaconda on https://anaconda.statuspage.io/incidents/qfv2n662hyym

[jakirkham]

Thank you both! 🙏