[DigitalPoint] App for Cloudflare®

[DigitalPoint] App for Cloudflare® 1.8.2

No permission to download
Why so strict? I am sure many people don't know it.
Why not help? Have corrected the code above, it is:
Because this thread has literally hundreds of users subscribed to it and they probably don't want to be getting push notifications/alerts/emails every time someone has a random Cloudflare question unrelated to the addon this thread is for. I'm happy to help with Cloudflare account questions when I can, but not in this thread. This thread is already busy enough, so best to keep it related to the addon.
 
Not part of the addon (yet?) but if anyone has XF Media Gallery, here are typical URLs for images.

Thumb:
https://mybucket.xxxxx.com/xfmg/thumbnail/0/49-8b7f3cb6dce3c196ecb1b306762c608c.jpg?1693938397

Full size:
https://mybucket.xxxxx.com/media/my-old-system.49/full?d=1698780922

I added both manually to Cache Rules. I don't know if one addon can detect the presence of another (that's above my pay grade and IQ level😕), but having these added would probably help others who aren't comfortable navigating Cloudflare.

Note that I have both of these served from my R2 buckets.
 
Well also, Cloudflare will cache .jpg files and adhere to whatever cache control comes from your web server. So for the public area with the thumbnails, normally it should cache that without any config or special rules. But if you definitely would need a rule for the private area where you are wanting to bypass XenForo permissions.
 
Sorry to trouble you @digitalpoint - but any chance you know what this means/how to resolve in forum/admin.php?cloudflare/ (when you click settings);

View attachment 293416



View attachment 293415

That probably has something to do with the power outage Cloudflare suffered yesterday. Some things still appear to be down. You can read more about it, including info from @digitalpoint, starting here:

 
That probably has something to do with the power outage Cloudflare suffered yesterday. Some things still appear to be down. You can read more about it, including info from @digitalpoint, starting here:


Thanks!
 
Found this out myself too...

Can it be that the Cloudflare outage throws CSP errors? Got a lot of them in my browser console...
 
Found this out myself too...

Can it be that the Cloudflare outage throws CSP errors? Got a lot of them in my browser console...
Maybe on their website, but doubtful if it’s your website. The main thing being impacted is you can’t change your Cloudflare settings/config reliably.
 
digitalpoint updated [DigitalPoint] App for Cloudflare® with a new update entry:

Better handling of unexpected Cloudflare API errors/problems

  • Added link for info about why each Cloudflare token permission is needed
  • Updated deep links into R2 buckets to use new URL endpoint
  • Suppress Cloudflare rate limit error when purging URLs from cache when guest page caching is enabled (a very high traffic site could hit API rate limits if there's a zillion posts flowing in at once)
  • Better handling of situation where Cloudflare API is down/unavailable
  • Cloudflare Workers that are created for the image proxy and unfurl proxy...

Read the rest of this update entry...
 
  • Updated deep links into R2 buckets to use new URL endpoint
  • Cloudflare Workers that are created for the image proxy and unfurl proxy have been rewritten to be ES Modules instead of Service Workers
Is there anything we would need to change on our end if we're already using R2 and cloudflare workers + image proxy?
 
Is there anything we would need to change on our end if we're already using R2 and cloudflare workers + image proxy?
The deep links are these links to the cloudflare dashboard, Cloudflare just updated the URL so they didn't exactly work correctly until digitalpoint fixed them just now.

1699074581987.png

And the workers update just uses cloudflare's new recommended ES module style of code instead of their older service worker style. Cloudflare claims the module style to be more efficient, but it probably isn't too significant.

As for getting it to update, I had to disable the image proxy in the XF App, then go to my workers in Cloudflare dash, delete the worker for the image proxy, and then go back into the XF App and re-enable the image proxy. That automatically updated the url proxy worker as well.
 
Is there anything we would need to change on our end if we're already using R2 and cloudflare workers + image proxy?
It's not really necessary. It's more just for new setups. But if you want to move to the ES Module style code, you can do as mentioned above (disable them, delete the Workers in Cloudflare and more or less start new and it will offer to create the Workers for you). But honestly, not really worth the effort. Not really going to make any difference whatsoever imo.
 
Details have arrived!


For some weird reason Cloudflare links do not unfurl on my board. They do here. Testing shows Could not fetch metadata from URL with error: Response returned a non-successful status code: 403. And I am using this add-ons proxy which uses Cloudflare 🤣
 
On the Cloudflare dashboard -> Workers & Pages -> Overview, I noticed that I have 3 proxies there. I guess two were automatically created (xf-url-proxy, xf-image-proxy), and at some point I created the 3rd one (xfimageproxy). I can't find info about it on the Overview or FAQ pages of this addon.

For xfimageproxy.admin-2ef.workers.dev, on the "Triggers" tab it has a custom domain xfimgproxy.mysite.com. Its DNS record is:
Type Worker
Name xfimgproxy.mysite.com
Target xfimageproxy
Proxy status Proxied
And the page xfimgproxy.mysite.com loads with a "hello world" message.

I searched this thread for "custom domain" but didn't find much. I'm guessing I can delete the custom domain and all 3 workers and then re-enable the image proxy setting in the XF app?

The only other related DNS record I see is one that was automatically generated by R2:
Type R2
Name data.forum.mysite.com
Target myforum-data
Proxy status Proxied
 
Details have arrived!


For some weird reason Cloudflare links do not unfurl on my board. They do here. Testing shows Could not fetch metadata from URL with error: Response returned a non-successful status code: 403. And I am using this add-ons proxy which uses Cloudflare 🤣
403 is forbidden... not sure why they would be blocking it, but maybe they are for whatever reason.

Using the Worker-based unfurl is working for me though when I test it:

1699080520027.png

On the Cloudflare dashboard -> Workers & Pages -> Overview, I noticed that I have 3 proxies there. I guess two were automatically created (xf-url-proxy, xf-image-proxy), and at some point I created the 3rd one (xfimageproxy). I can't find info about it on the Overview or FAQ pages of this addon.

For xfimageproxy.admin-2ef.workers.dev, on the "Triggers" tab it has a custom domain xfimgproxy.mysite.com. Its DNS record is:

And the page xfimgproxy.mysite.com loads with a "hello world" message.

I searched this thread for "custom domain" but didn't find much. I'm guessing I can delete the custom domain and all 3 workers and then re-enable the image proxy setting in the XF app?

The only other related DNS record I see is one that was automatically generated by R2:
xfimageproxy was never created by this addon. You can grep for that string and it doesn't exist in the addon, so no clue where that came from... probably someone made it manually? If it's a "Hello World" message, it's just a default Worker with the default code it plops in when you manually create a Worker. If you aren't using it for something, it should be safe to delete it. It has nothing to do with this addon either way.

You can delete them all and recreate if you'd like, but if they are working already, no real point... just leave it "as is". The two Workers that the addon creates do not get assigned a custom domain route or anything (the URLs are never seen by end users, the Workers are just used on the backend, (so no need to assign "pretty" public URLs to them).
 
Top Bottom