I also encountered the same problem when I was helping my friends in the ecosystem to deploy. I hope it can be solved soon.
The deployment was normal before, and it was also normal to access it through canister_id plus the domain name of icp0.io.
The request_id is returned normally when registering the domain name in the boundary nodes. But when using this request_id to query the registration status, it always returns “failed to create dns record”.
I don’t know if this is related to the number of boundary nodes and the overall service performance, because it’s not just this problem. I also encountered node congestion when deploying a few days ago.
I am happy to report that the custom domain service has been restored. Registrations work again. Registrations that have been submitted before the fix will be retried every few minutes and should go through soon. I am very sorry for the inconveniences caused!
There seems to be an issue in the docs page. When you click on the copy link of a code box, some of the whitespaces, are actually U+40, so called “no break” spaces. This is a bug of the website and I will forward it to the corresponding team.
You can verify it yourself by pasting it, for example, on this website and examining it.
I have the same problem with the alias, just 4 random IPs… Very strange
Maybe it takes time for NameCheap to update the Alias, but I’ve been waiting almost whole day…
I would wait until the BN team responds. I forwarded the info. From my experience, the more you hack around DNS without exactly knowing where the issue lies, the worse it gets. Just wait for feedback.
Sorry for the issues! It was a problem on our end and is fixed now. Since the service retries several times, @jamesbeadle your registration went through now after the fix.
@AlexEmets please check again the status of the registration. If didn’t go through (i.e., status is available), let me know.
Quick explanation what happened: Cloudflare (provider that we use for the DNS-01 ACME challenge of let’s encrypt) deprecated some fields in their API, but did not update their rust library accordingly. Hence, the library started failing to interact with the Cloudflare API. We created a manual patch and everything is working again.