Cache is 512MB and you dont have to cache files you can just serve it without cache 🤷

Is it possible to have one subdomain for all R2 buckets? It seems to only want to connect a domain to one bucket but that's an issue for me

Not currently, no

I noticed the URL for my buckets was https://XXXXXXXXXXX.r2.cloudflarestorage.com/bucketname

Will the xxxxxxxxxx ever change? If so, can I use a service that does CNAME aliasing to do it outside of cloudflare?

Nope and nope.

The xxx is your account ID and you can't CNAME to CF w/o using SSL for SaaS, which you can't setup yourself for R2.

blegh

thanks

Thanks for your response, but I've got that effectively in Github right now. Given that is prone to fall over at different times of the day, my thinking is that a CDN delivered exe could alleviate those issue. I keep reading bits and pieces about a Cloudflare CDN, but I don't see it listed as a service specifically - I assume this is just referring to the cache solution?

Unlike other Cloud providers, Cloudflare’s CDN isn’t a single product, it is all of CF’s products combined. There’s WAF to reduce requests to your origin, Rules to rewrite requests on the fly, Workers for Compututation at the Edge, R2 for zero egress fees, Argo for speeding up origin connections, the list goes on

And all of those can be considered the CF CDN, since they help you “deliver content”

Cache is one of those products that can make your life easier

I see - so all the services replicated across all the servers, most of the time sort of scenario? Thanks for the explanation

Not always.

For example, you can define regional services for things like GDPR

Sorry for these painful questions - still trying to get to grips with it all, and I'm a little slow 😉

Or R2, which is currently just a single region, but should be replicated/auto-migrated soon

The cool thing with R2 is that you pay per request, not for bandwidth

ok - so would R2 (after the replicated scenario comes into play) be right to use as a quasi CDN to serve the exe through a worker?

I think that is where I'm getting a little confused, with regard to the storage of the object sizes in different services, versus the caching of those

You can use R2 right now to serve an EXE and you can use a public url instead of needing a worker at all 🤷

I mean, it works as a “CDN” now, it may just become faster later on

And for an EXE, the latency isn’t nearly as sensitive as other applications might be

Yeah, I'm not sure why that was not clicking in my head - I'm already doing that with content, so why not an exe

Perfect

ok - thanks @HardAtWork & @Unsmart | Tech debt - you've both cemented something into my head. I need a cup of tea about now. Cheers for everything

Ah that makes sense. I read the docs, but I couldn't figure out if something like that was going on. Thanks 👍

Woohoo got it working

Holy crap though, I thought doing S3 over curl wouldn't be that bad, obviously I was wrong