Unable to get Recording URL's filename and extension using Python

[abhilash]

In the API Cloud Recordings API, when I try to use Python’s urlopen or requests to fetch the filename and extension of the download_url field, it gives me 403 Forbidden error.

I’m using the download_url of the recording that has type active_speaker

How should I fix this?

[tommy (Tommy Gaessler)]

Hey @abhilash,

Do you know if your Cloud Recording settings are password protected or private?

Check out my post here if so,

BUG downloading cloud recordings with access token set results in an invalid response

@patricio.giacomino, @matt, @ryan As of now to download a private or password protected cloud recording you must use a Zoom JWT App Type. Use the download_url from the Get Recordings Endpoint or the download_url from the Recording Completed Webhook and add the access_token query param to the end of the url with your JSON Web Token as the value. Example: https://api.zoom.us/recording/download/{{ Download Path }}?access_token={{ JWT Token }} [19%20PM] The only way to use OAuth to download a …

Let me know if this helps!

Thanks,
Tommy

[abhilash]

It’s already set to off.

[tommy (Tommy Gaessler)]

Thanks @abhilash,

What app type are you using: OAuth or JWT?

[abhilash]

I’m using OAuth…

[ryan]

We’ve started seeing 403s on all of our recording download requests as of yesterday. Something must have changed.

[abhilash]

Oh, ok. So, what’s next? Any ETA until this problem is resolved?

[ryan]

This was not correct, see below

Doing some more investigating… We previously used the download_token in the recording.completed webhook payload as an access_token param on the recording download url… now it appears that omitting the access_token param returns a 200 but adding it returns the 403…

[abhilash]

Meanwhile, @tommy and @ryan Thanks a ton for replying so fast! I really appreciate your responsiveness.

[ryan]

I’m fast because my customers are complaining about not being able to access their recordings!

[ryan]

I was wrong about download_token… some user agents (httpie, browsers) are able to fetch the recordings with the download token i.e. https://zoom.us/recording/download/<recording path>?access_token=<download_token>. But others are not (curl). Still investigating…

[ryan]

Ok, that should have tipped it off! It’s cookies! The initial request returns a 302 and sets some cookies in the response. You must use those cookies when following the redirect.

[TimZoom (Tim)]

Thanks for the info Ryan, we are escalating to our eng team and will be back with a response asap.

[matt]

Can the status page be updated? This is a service disruption.

[TimZoom (Tim)]

We don’t have a status page for the API/Developer tools yet. We plan to release one later this year though to be more transparent about these sorts of issues.

[drosentrater (David)]

I’m having the same issues using a webhook only app. Download urls were working fine until Friday and then started sending back errors. Thanks for being on top of this!

[tommy (Tommy Gaessler)]

Thanks for posting @drosentrater, our team is working on the issue and we will post back here with updates.

Thanks,
Tommy

[proveedores (Softdynamic)]

¿Se llegó a solucionar este problema ?
Porque me sucede lo mismo con download_url

[eric]

FYI, this is no longer working for me either in Java. I think you are correct about the cookies not being used on the redirect by the library I am using to download the recording.

[michael_p.zoom (Michael P.)]

Hi Eric,

For updates, please refer to our this forum post - Update: Recording download url 403 error. We’re working on resolving the issue asap.