DI have a fallback rule at the bottom so it won't fail to redirect
D
Jwhy did my random website that usually has no traffic have 12000 requests today and now it doesn't work?
Csounds like it got attacked?
Jbut why tho
it doesn't do anything important, it's just a google site
it doesn't do anything important, it's just a google site
J
Cyou'd have to check on your origin for why it isn't working now though, if the ddos is over can't mitigate it in the past, perhaps exceeded some resource limit though
Cnew iata codes/pops? I was looking over my data and maybe 2 a month or so and then nothing for a few months
There's new colos at the time though, mutiple times a month. Sometimes they might just being pulled/renumbered though
There's new colos at the time though, mutiple times a month. Sometimes they might just being pulled/renumbered though
Dcool, I won't worry about it too much then. I have cache everything enabled anyway, so even the wrong region wouldn't be that bad.
DI'm almost sure the answer is no, is there any way I could use the same hostname for multiple public R2 buckets, without a Worker, snippet, or load balancing?
Cit uses cf for saas under the hood, greedily grabbing control of the hostname
Dwhat does?
Cr2 custom domains
Dstupid question, is there any weirdness when you do a fetch from a Worker in the same zone as the r2 custom domain?
CR2 Custom Domains are workers under the hood, but just like Pages, CF for SaaS exempts you from the typical worker-worker same-zone weirdness
Cthere's no advantage I know of using R2 Custom Domains over bindings though
Ceitherway all you get is local cache
Ccan use something like kotx/render to take care of caching, range headers, etc
Dwon't you always get billed when using the bindings, even if it's in the cache?
Cfor the worker? sure
Same for if worker fetches r2 custom domain tho
Same for if worker fetches r2 custom domain tho
Dwait, what :/
CWorkers run in front of cache
Dif you do:
Won't that use the cache that's on the r2 custom domain?
Won't that use the cache that's on the r2 custom domain?
CWorker using R2 Binding with Cache API or kotx/render is going to charge you class b if no cache hit, otherwise just worker
Worker using R2 Custom Domain with caching enabled is going to cost you class b if no cache hit, otherwise just worker
eitherway no tiered cache, not supported by r2 custom domains
Worker using R2 Custom Domain with caching enabled is going to cost you class b if no cache hit, otherwise just worker
eitherway no tiered cache, not supported by r2 custom domains
Dand a sanity check, if you visit
https://azure-di-source-files.auto.ai.moda/83d6300e2bb85d71b49bedd61f19fc37?cache-control=private%2C+immutable%2C+no-transform%2C+max-age=31536000%2C+stale-while-revalidate=31536000%2C+stale-if-error=31536000%2C+s-maxage=0&expires=Thu%2C+01+Jan+1970+00:00:00+GMT&ttl=604800&verify=1711929600-AXsEtplmiIONuUwUTfXh%2BVbEtqDo4rE0kPjgyN37j6w= in your browser right now, I don't get charged a class b, right?Conly if its cached
Dassuming it's cached, no charge for a class b?
Cright, you're still paying for the worker request to get you redirected there though
Cunless you're using transform rules
Cwhich I guess you are, opps
DI am, remember the messed up thing we just did lol
Dhow does cloudflare tunnel work with multiple connectors? does it automatically load balance?
CI may have forgotten about it in the 4 seconds after
CIf you're involving a worker it's the same cost if using r2 custom domains fetch or bindings
Cbut yea you can avoid the worker cost entirely by using the r2 custom domain direclty
Dand both use the same amount of caching?
Cright, r2 custom domains don't supported tier caching so you just get local
Cand cache api is just local as well
CBy design, replicas do not offer any level of traffic steering (random, hash, or round-robin). Instead, when a request arrives to Cloudflare, it will be forwarded to the replica that is geographically closest. If that distance calculation is unsuccessful or the connection fails, we will retry others, but there is no guarantee about which connection is chosen.https://developers.cloudflare.com/cloudflare-one/connections/connect-networks/deploy-tunnels/deploy-cloudflared-replicas/
Dis
cacheEverything: yes exactly the same as this?
Cthere might be some minor behavior differences but the outcome is mostly the same, still with r2 just going to get local colo cache, up to the ttl you specify
Dwell so in my case, I needed to enable eligible for cache for the PDFs to be cached. Without it, they weren't being cached iirc
Cthe one advantage cache api has in workers is that you can specify any ttl, even lower then the edge ttl min for your plan
Cnot worth it to pay the worker invocation costs just for that though I'd imagine lol
Dwhenever possible, I try to make S3 objects immutable.
JI bought the domain with Google domains (rip) and I don't know much about DDoS attacks or web servers.
D@Chaika another stupid question, do cache rules do anything ever if there's a Worker already listening on that custom domain?
Dnothing right?
CI mean the worker itself won't ever be cached, but they would effect the origin behind it (if any), iirc if you're fetching a non-cf zone it would also use your cache settings
Dah right, I remember now that we did that to enable caching on Backblaze B2 for my Workers.
