@Rishabh Jalan R2 _should_ be faster, from what I understand it's globally distributed, as in it sho
@Rishabh Jalan R2 should be faster, from what I understand it's globally distributed, as in it should perform like CloudFront + S3 automatically
Rs3 vs r2 in putting string 'helloworld' r2 put takes upward of 1.5 seconds where as s3 put takes around 0.2 seconds
get request r2 takes upward of 0.5 seconds , s3 takes around 0.05
get request r2 takes upward of 0.5 seconds , s3 takes around 0.05
RI shifted under the same assumption but its a lot different
Sthat is concerning, but I think R2 is optimized for thoughput, not latency, like large media assets
Rhow to optimise this latency ? any idea ?
Smaybe something changed though, like I mentioned, R2 was doing great for about 2 months, but since beginning of April it's been nothing but problems
Rhmm , i may need to go back if its this slow . its like 10x slower
CIt's not distributed at all except for cache/the fact you could manually make buckets in a bunch of regions. Under your r2 bucket settings you can see the location of your bucket. All your objects stay in that region, only caching would let you distribute them globally
Rmy r2 bucket is in apac region and i am calling it from apac only still very slow
Zapac is known to be unstable
RCan anyone check and let me know what is the latency ?
Rdropping the python script
s3vsr2speedtest.py1.41KB
R@Space - Ping in replies thank you so much . its better than apac but still not as good as s3
Rhas the quality deteriorated recently ?
Ris there any other alternative ? else i have to go back to aws s3
Rlet me check out buffer writes . never heard about it
Ri am using aws lambda to edit some images . any idea on how i can buffer writes ?
Rhow is workers latency in apac ?
Rthat i know but how do they perform vis aws lambda
Ris queues similar to sqs?
Rsqs is simple queue service from aws
Rsounds like a lot of work to move from aws lambda to workers .
Rlambda bill is around 10 dollars only as of now
Ri will move when this increases right now s3 has started reaching 40
Rso moving from that
Rhmm ?
R40 includes egress
Ri am expecting it to rise slowly so taking appropriate actions
Rthanks for your help @Space - Ping in replies
Rthats correct but egress later is cheaper
Rshould balance it out
S@Space - Ping in replies when I mentioned earlier about R2 being globally distributed, I was wrong as you mentioned, but if you proxy it as a public bucket, it sits behind cloudflare's caching capabilities
Sthat's my use case at least, I don't really see why you wouldn't do that if R2 is meant for public assets
Smy current bucket is behind ENAM (eastern north america), I'm gonna try WNAM and see if things improve
Soh interesting, that's good to know
Syou're right, the cache proxying may actually be the problem
Sya, that's what attracted R2 to me, they seem to be a great choice if you're hosting a package repository with lots of read traffic
Rokay one more noob question , can i cache s3 using cloudflare in some way ?
Sya, it's two things that make people put up with the quirks, cost savings and better performance for your use case
RHmm correct i will do this. This is the best
Chello, i added an R2 to test and I added 2 domains (subdomains of domains on cloudflare dns).. the first worked fine, but the second doesnt resolve, i waited a while, i also tried removing it and readding it .. does anyone have any suggestions ?
EWhats the domain? Do you see a DNS entry in the DNS settings?
Cyes i see a dns entry for it - pictures-test.serranoaustralia.com.au
C
SIt's working for me
S
Cyes its working for me now
Ci guess i had to wait a bit longer
Ckids these days, so impaient 
thank you for the help !
thank you for the help !
MI need to send files to my bucket, files larger than 300mb, any suggestions? I've already tried with workers and S3 api and nothing works
