Account ID and token ID (not the actual secret) would be great. I also assume it’s midnight UTC (har

MMatt Silverlock Account ID and token ID (not the actual secret) would be great. I also assume it...

K

kev-ac•1/9/23, 2:38 PM

Just sent you the details in a PM. Timestamp according to the Audit Log was 2023-01-09T01:00:57+01:00.

JJeff Is there a way to get the folder size?

V

Vero•1/9/23, 4:11 PM

Hey @j3fftw . Currently, there's not any great way to do that beyond listing all the files which start with the same prefix and summing the file size of each one

MMatt Silverlock Account ID and token ID (not the actual secret) would be great. I also assume it...

K

kev-ac•1/10/23, 9:15 AM

Followed up with you in PM, but wanted to update the channel here too, just in case somebody else finds this later. It happened again this midnight for a 24-hour token.

E

Epailes•1/10/23, 1:19 PM

I tried having a search around but am I right in saying that there's no replication for R2 at the moment, but some form of redundancy/replication is on the roadmap?

S

sakuta•1/10/23, 4:49 PM

how can i allow only access to a specific bucket for an api token? (using s3 api, not workers binding)

S

sakuta•1/10/23, 4:49 PM

and not access to all buckets in my account

K

kian•1/10/23, 4:52 PM

Not possible currently

K

kian•1/10/23, 5:20 PM

- await env.PROFILE.put('do-something', file);
+ await env.PROFILE.put('do-something', file.stream());

- await env.PROFILE.put('do-something', file);
+ await env.PROFILE.put('do-something', file.stream());

K

kian•1/10/23, 5:23 PM

Depending how the uploads are taking place, it's preferred to send the file's binary content as-is in the request body & stream that into R2 so that you're not buffering the file into the Worker's memory which can cause resource exhaustion if there's large enough files in concurrent requests - but if you're using something like a HTML

<form>

<form>

then you'll probably have to stick with FormData.

K

kian•1/10/23, 5:23 PM

.put

.put

's third argument, options, is how you can specify

content-type

content-type

and whatnot to be stored with the object so that they're properly presented when viewed in a public bucket and the like.

K

kian•1/10/23, 5:28 PM

Basically

--data-binary

--data-binary

in cURL or just

binary

binary

in Postman's body options.

interface Environment {
    R2: R2Bucket;
}

export default <ExportedHandler<Environment>>{
    async fetch(req, env) {
        if (req.method !== 'POST') {
            return new Response('Method Not Allowed', {
                status: 405,
                headers: {
                    Allow: 'POST',
                },
            });
        }

        const url = new URL(req.url);
        const path = url.pathname.slice(1);

        try {
            await env.R2.put(path, req.body, {
                httpMetadata: req.headers,
            });

            return new Response(`Successfully uploaded ${path}`);
        } catch (e) {
            return new Response(`Failed to upload ${path}`, {
                status: 500,
            });
        }
    },
};

interface Environment {
    R2: R2Bucket;
}

export default <ExportedHandler<Environment>>{
    async fetch(req, env) {
        if (req.method !== 'POST') {
            return new Response('Method Not Allowed', {
                status: 405,
                headers: {
                    Allow: 'POST',
                },
            });
        }

        const url = new URL(req.url);
        const path = url.pathname.slice(1);

        try {
            await env.R2.put(path, req.body, {
                httpMetadata: req.headers,
            });

            return new Response(`Successfully uploaded ${path}`);
        } catch (e) {
            return new Response(`Failed to upload ${path}`, {
                status: 500,
            });
        }
    },
};

interface Environment {
    R2: R2Bucket;
}

export default <ExportedHandler<Environment>>{
    async fetch(req, env) {
        if (req.method !== 'POST') {
            return new Response('Method Not Allowed', {
                status: 405,
                headers: {
                    Allow: 'POST',
                },
            });
        }

        const url = new URL(req.url);
        const path = url.pathname.slice(1);

        try {
            await env.R2.put(path, req.body, {
                httpMetadata: req.headers,
            });

            return new Response(`Successfully uploaded ${path}`);
        } catch (e) {
            return new Response(`Failed to upload ${path}`, {
                status: 500,
            });
        }
    },
};

interface Environment {
    R2: R2Bucket;
}

export default <ExportedHandler<Environment>>{
    async fetch(req, env) {
        if (req.method !== 'POST') {
            return new Response('Method Not Allowed', {
                status: 405,
                headers: {
                    Allow: 'POST',
                },
            });
        }

        const url = new URL(req.url);
        const path = url.pathname.slice(1);

        try {
            await env.R2.put(path, req.body, {
                httpMetadata: req.headers,
            });

            return new Response(`Successfully uploaded ${path}`);
        } catch (e) {
            return new Response(`Failed to upload ${path}`, {
                status: 500,
            });
        }
    },
};

K

kian•1/10/23, 5:29 PM

That'll take a binary sent to

https://example.com/dog.png

https://example.com/dog.png

and upload it to R2 as

dog.png

dog.png

K

kian•1/10/23, 5:30 PM

I'm not sure I follow what you mean by over JSON

K

kian•1/10/23, 5:31 PM

You'd just send the file's raw content as-is as the body

K

kian•1/10/23, 5:51 PM

When you setup a custom domain for a public bucket then you'll get caching and all the other Cloudflare features

K

Karew•1/11/23, 12:17 AM

It depends who you want to process the videos. If you want Cloudflare to process the videos, use Stream. If you want to process the videos and manage the files, use R2. You'll probably also still need to use R2 for things like still images, avatars, etc

K

Khagan•1/11/23, 5:05 AM

I am getting "CORS error". How can i configure CORS settings on R2?

K

Karew•1/11/23, 5:08 AM

https://kian.org.uk/configuring-cors-on-cloudflare-r2/

K

Khagan•1/11/23, 5:21 AM

Does it look good? I posted this but still getting the same error.

K

Khagan•1/11/23, 5:34 AM

HTTP GET Output;
<?xml version="1.0" encoding="UTF-8"?><CORSConfiguration xmlns="http://s3.amazonaws.com/doc/2006-03-01/">
<CORSRule>
<AllowedMethod>GET</AllowedMethod><AllowedMethod>HEAD</AllowedMethod>
<AllowedOrigin>*</AllowedOrigin><AllowedHeader>Authorization</AllowedHeader><MaxAgeSeconds>3000</MaxAgeSeconds>
</CORSRule>
</CORSConfiguration>

HTTP GET Output;
<?xml version="1.0" encoding="UTF-8"?><CORSConfiguration xmlns="http://s3.amazonaws.com/doc/2006-03-01/">
<CORSRule>
<AllowedMethod>GET</AllowedMethod><AllowedMethod>HEAD</AllowedMethod>
<AllowedOrigin>*</AllowedOrigin><AllowedHeader>Authorization</AllowedHeader><MaxAgeSeconds>3000</MaxAgeSeconds>
</CORSRule>
</CORSConfiguration>

HTTP GET Output;
<?xml version="1.0" encoding="UTF-8"?><CORSConfiguration xmlns="http://s3.amazonaws.com/doc/2006-03-01/">
<CORSRule>
<AllowedMethod>GET</AllowedMethod><AllowedMethod>HEAD</AllowedMethod>
<AllowedOrigin>*</AllowedOrigin><AllowedHeader>Authorization</AllowedHeader><MaxAgeSeconds>3000</MaxAgeSeconds>
</CORSRule>
</CORSConfiguration>

HTTP GET Output;
<?xml version="1.0" encoding="UTF-8"?><CORSConfiguration xmlns="http://s3.amazonaws.com/doc/2006-03-01/">
<CORSRule>
<AllowedMethod>GET</AllowedMethod><AllowedMethod>HEAD</AllowedMethod>
<AllowedOrigin>*</AllowedOrigin><AllowedHeader>Authorization</AllowedHeader><MaxAgeSeconds>3000</MaxAgeSeconds>
</CORSRule>
</CORSConfiguration>

KKarew <https://kian.org.uk/configuring-cors-on-cloudflare-r2/>

K

Khagan•1/11/23, 5:40 AM

It loaded after a few minutes. Thank you!

Uu007 is there a s3 compliant tool to transfer r2 objects from 1 account to another?

K

Karew•1/11/23, 6:45 AM

Rclone is awesome https://rclone.org

K

kian•1/11/23, 6:46 AM

Yeah,

rclone

rclone

is the way to go - CopyObject can't do cross-account so that's ruled out.

K

Karew•1/11/23, 7:43 AM

You can also copy between two remote accounts

K

Karew•1/11/23, 7:43 AM

It uses the computer running rclone as an intermediary

K

Karew•1/12/23, 9:42 PM

Page Rules are the old system, use Cache Rules

E

Erisa•1/12/23, 9:43 PM

page rules are the old way, cache rules are the new way
both will work functionally the same

U

Unsmart•1/12/23, 9:56 PM

page rules will eventually go away afaik so you should use cache rules

E

Erisa•1/12/23, 9:57 PM

tbh yeah its good to get used to cache rules earlier rather than later

U

Unsmart•1/12/23, 10:03 PM

Not sure if they have any documentation about how they evict stuff but if I had to assume its a very basic least used gets thrown out first with some added multiplier for the plan level you are on.

They will respect the edge cache ttl as far as not storing it longer than what you set but it can get evicted before that time is reached

Sskaboy007 so no practical differences

K

Karew•1/12/23, 10:04 PM

The new rules are way more powerful. You can construct really complex expression and check headers/IPs/requests and stuff to see if rules apply. Page Rules only pattern match on the URL

Account ID and token ID (not the actual secret) would be great. I also assume it’s midnight UTC (har

Similar Threads