Hello,
I am experiencing what appears to be a Vercel Anycast/BGP routing or network reachability issue from my internet connection in Oman.
Multiple unrelated custom domains hosted on Vercel are timing out from the same network, while their corresponding .vercel.app domains work normally.
Affected examples include, but are not limited to:
safetyhero.omnexus.looksintentional.comforlio.looksintentional.comuat-nexus.looksintentional.compos.london.edu.omkayan.omelbrusagency.comThis is not limited to one project or one root domain.
Current behaviour
The affected custom domains ultimately resolve to the following Vercel IP address:
216.198.79.1Connections to TCP port 443 on this IP time out from the affected Oman network.
Example:
curl.exe -I -v --connect-timeout 15 ` --resolve safetyhero.om:443:216.198.79.1 ` https://safetyhero.om/Result:
Trying 216.198.79.1:443...Connection timed out after 15003 millisecondscurl: (28) Connection timed outHowever, when I force the same hostname to another Vercel IP, the connection succeeds:
curl.exe -I -v --connect-timeout 15 ` --resolve safetyhero.om:443:64.29.17.1 ` https://safetyhero.om/With 64.29.17.1, the TLS connection succeeds and Vercel returns a valid HTTP response for the same hostname.
The corresponding .vercel.app domains are also accessible normally from the same network.
DNS configuration
The DNS configuration is valid and does not use manually hardcoded Vercel IP addresses.
For example, looksintentional.com is managed through AWS Route 53, and each Vercel subdomain uses the project-specific CNAME supplied by Vercel:
nexus.looksintentional.com CNAME → project-specific vercel-dns hostnameforlio.looksintentional.com CNAME → project-specific vercel-dns hostnameuat-nexus.looksintentional.com CNAME → project-specific vercel-dns hostnameVercel shows these domains as:
Valid ConfigurationTherefore, this does not appear to be an incorrect DNS record or project-domain configuration issue.
Test summary
216.198.79.1:443 → connection timeout64.29.17.1:443 → connects successfullyvercel.app → works normallycustom domains → time out when routed to 216.198.79.1Because several unrelated Vercel projects and domains are affected, this appears to be an infrastructure-level routing or reachability issue rather than an individual project problem.
The test results suggest a possible routing issue affecting:
216.198.79.0/24from the affected Oman ISP/network or an upstream network path.
Diagnostic output
I ran Vercel’s Windows connectivity diagnostic script from the affected network:
Invoke-RestMethod -Uri https://raw.githubusercontent.com/vercel-support/vercel-connect-debug/main/vercel-debug.ps1 |Invoke-Expression |tee vercel-debug.txtThe complete diagnostic output is available here:
https://gist.github.com/looksintentional/6982be532579db40d528da332fad4b99
The diagnostic was generated directly from the affected network in Oman.
I can also provide additional traceroutes, DNS results, screenshots, or tests from other Oman networks if required.
Could someone from the Vercel networking or infrastructure team please investigate reachability to 216.198.79.0/24 from Oman?
I would also appreciate confirmation on whether there is a supported temporary workaround for affected production custom domains. I do not want to replace the project-specific Vercel CNAME records with an undocumented IP address.
Thank you.
