I'm getting ERR_CONTENT_DECODING_FAILED on r2 public file - It is common issue? When I try to search
I'm getting ERR_CONTENT_DECODING_FAILED on r2 public file - It is common issue? When I try to search in discussions I find results but no solution.
CIsn't it a network problem of R2?
Sthanks, I have tried multiple times, also same file renamed to .txt works :/
CActually the program will retry 3 times, but unfortunately all of them fail
S
resolved by removing ContentEncoding header via CyberDuck on uploaded file.
PSomething like this?(copilot)
Jfacing very low speed in r2 using custom domain, public domain and also with worker. Users are complaining about speed any way to fix it?
JYes I was using cache previously but after checking your old replies in this thread I have disabled the workers cache ( caches.default ) and also added a cache bypass rule for particular subdomain which is connected to worker
Will see in 24 hours if there is a major difference in performance
currently after removing the cache, it's working good ( not so fast but adjustable)
Will see in 24 hours if there is a major difference in performance
currently after removing the cache, it's working good ( not so fast but adjustable)
JI think cache is also affecting smaller files as I am on free cloudflare plan , and in paid worker plan
JI can use a r2 custom domain but then anyone can leech my content easily. To prevent it I am using a worker with a custom subdomain so that I can calculate some verification steps
JI was thinking of using some base64 decode function which is available in the free zone , will obfuscate the base64 string with some random prefix and can use substring function in WAF rule
Ji haven't tested it yet
Jyea 
I know it but after obfuscating it can confuse the leecher a little not much
I know it but after obfuscating it can confuse the leecher a little not muchJYou are right, Thanks for your help
KHi,
Files uploaded to r2 bucket are not being instantly available with public domain. I need to wait 2-3 minutes to get it, which is not looking ok for me. Are there any ways to fix it?
Files uploaded to r2 bucket are not being instantly available with public domain. I need to wait 2-3 minutes to get it, which is not looking ok for me. Are there any ways to fix it?
KYes
KA first request is being made for an existing file, so cache theoretically should not be an issue
Also I've noticed that the problem getting worse when I'm uploading files concurrently
Also I've noticed that the problem getting worse when I'm uploading files concurrently
KAdded rules to bypass cache for the whole subdomain. Didn't help 
AHas anyone successfully implemented service worker to serve partial content like mp4 ?
Dwhy no rest api for r2
Du mean using s3 compatibility api?
Dman its compplex af
Di just want something like curl PUT mybucket/myfile token=sometoken
Dthen need wrangler
Dwrangeler need node
Di just want to avoid bloatware
Ds3 api will do
BHi there!
I'm running a Docker container on a VPS. I want to write a shell script that can
put files from the container into a Cloudflare R2 bucket. I've read through the
docs, and it seems like my best options are: wrangler, rclone, aws CLI, Workers API.
Is it not possible to simply use a curl command to a Cloudflare REST API
endpoint? For example:
I'd prefer to do something like this, rather than add
to my Docker container, or spin up a Cloudflare Worker. Thanks in advance
for the help!
I'm running a Docker container on a VPS. I want to write a shell script that can
put files from the container into a Cloudflare R2 bucket. I've read through the
docs, and it seems like my best options are: wrangler, rclone, aws CLI, Workers API.
Is it not possible to simply use a curl command to a Cloudflare REST API
endpoint? For example:
I'd prefer to do something like this, rather than add
wranglerrcloneto my Docker container, or spin up a Cloudflare Worker. Thanks in advance
for the help!
CThe r2.cloudflarestorage.com endpoints are just for S3.
There is an undocumented REST API that Wrangler uses, but it has limits like ~300 mb upload size, and it's slow as it has to go through Core/not processed at edge, plus it may change at any time and break since it's undocumented
Using one of the many s3 libs or rclone which use the s3 compat api would work well. You could write your own simple worker to use the Workers API, but would have to deal with multipart if files are bigger then 5 GB, which would be easier with a lib or rclone
There is an undocumented REST API that Wrangler uses, but it has limits like ~300 mb upload size, and it's slow as it has to go through Core/not processed at edge, plus it may change at any time and break since it's undocumented
Using one of the many s3 libs or rclone which use the s3 compat api would work well. You could write your own simple worker to use the Workers API, but would have to deal with multipart if files are bigger then 5 GB, which would be easier with a lib or rclone
BThanks Chaika for all the help! Is there anything you can point me to on the undocumented REST API that Wrangler uses? The limits on size and speed don't bother me. If not, I'll just use rclone 
Copps the limit on wrangler/rest is ~300 mb or so, not 500 mb
Cthere is no documentation for the undocumented api ;p
I explained it more to show why it's a bad idea to use wrangler as well, I would just use rclone.
I explained it more to show why it's a bad idea to use wrangler as well, I would just use rclone.
BHaha that makes sense! I'll just use rclone. Thanks again for all your help, I really appreciate it.
LIs their no way to skip time for big file?
Lvery bad
LWhere do I complain about this?
Lreally? you mean page rule?
VHey, while saving chunks for multipart uploads, I'm sporadically getting 502 and 503 errors. I suspect this might be related to request throttling, as the error rate spiked significantly after increasing my traffic from 2K to 10K/day upload. When I retry uploading a chunk, it usually succeeds. I've checked the documentation but couldn't find any relevant details. I came across some forum threads, like https://community.cloudflare.com/t/r2-intermittent-500-errors-on-put/485134, but none seem to offer a clear resolution. Does anyone know why these intermittent 502/503 errors are occurring and how I can resolve them?
Lhow? can you help me?
VI'm uploading to multiple buckets and don’t expect to upload 300 chunks to the same one. Does the 300 concurrent request limit apply across all my buckets collectively, or is it per bucket? In other words, if I send 300 concurrent requests spread across different buckets, would I still hit that limit?
Lok thanks
Vis there any other limit across buckets?
Lwow Thanks. It starts working magically.
NDoes anyone know, if I begin an upload then abort it, will the bytes end up in the bucket, or automatically cleaned up?
NI’ll test it myself later if not, just wondering
CMultipart uploads are aborted after 7 days by default (can modify in settings under lifecycle rules). If it's just a normal PUT that gets killed shouldn't go anywhere
NSorry I’m just talking about a PUT
NWhen you say shouldn’t go anywhere you mean the incomplete upload never makes it into the bucket?
IIf a regular PUT is aborted halfway through the data transferred is just discarded, it never goes anywhere (including the bucket)
NThanks