Cloudfront cache not cleared for database dumps with Accept-Encoding request headers

[LawnGnome]

Not trying to bump a closed issue thread, but I'm also not sure if this is related. I'm being served a stale version of the dump when I don't provide a Accept-Encoding header:

Setting Accept-Encoding to gzip, deflate, br grabs me the new data, but I'm actually not sure if that's just related to how cloudfront is caching the request.

Is there any way around cloudfront's cache or should I just expect to possibly have ~48-hour stale data?

Originally posted by @enricozb in rust-lang/crates.io#6652 (comment)

(opening per @jdno's suggestion)

[enricozb]

There were some questions about reproduction, and I can still reproduce this today:

In case it matters (for cache invalidation propagation or something) I'm in Honolulu Hawaii.

[jdno]

Thanks for clarifying, @enricozb! I might've confused the output from production and staging, I see the same as you when testing again. My bad.

[Turbo87]

@rustbot transfer infra-team

[rustbot]

Failed to transfer issue:
error: rustbot does not have the correct permissions to execute TransferIssue