Blog
How to put your status page on status.example.com or a separate domain: CNAME records, certificate issuance, independent DNS and the mistakes that stop it from resolving.
A custom domain status page setup takes four steps. Pick a hostname such as status.example.com, create a CNAME record pointing to the target your status page provider gives you, let the provider issue a TLS certificate once the record resolves, and verify everything with dig and curl. The record itself is rarely the problem. Most failures come from dependencies: if the status page shares DNS, hosting or a CDN with your product, it goes down at the same moment.
This guide covers the hostname decision, the exact records, how certificates get issued, which dependencies to cut, and the DNS mistakes that keep a status page from resolving on the day you need it.
There are two reasonable choices, and the trade-off is recognition against independence.
This is the default for good reasons. Customers recognize it, it looks official, and it sits next to your product in browser history and search results. A phishing-wary customer is more likely to trust status.example.com than an unfamiliar domain.
The catch is DNS. Every lookup for status.example.com first goes through the authoritative nameservers of example.com. If your DNS provider has an outage, or someone breaks the zone, the status subdomain stops resolving together with the product. Delegating the subdomain with its own NS records does not fix this, because the referral to those nameservers still comes from the parent zone.
A separate domain, registered at a different registrar and hosted on a different DNS provider, survives a failure of your main zone, an expired main domain, or a botched DNS migration. The cost is weaker recognition, so link to it prominently from your app, docs, support replies and error pages so customers learn it before an incident.
If your product and its status page share a DNS zone, they share an outage. The hostname is a reliability decision, not a branding one.
A practical middle ground: use status.example.com, but host example.com on two DNS providers (primary plus secondary) and keep a long TTL on the status record so caches carry it through short DNS failures.
dig +short status.example.com ANY should return nothing. A leftover A, AAAA or TXT record at the same name blocks the CNAME.dig +short example.com CAA. If any exist, confirm they allow the certificate authority your provider uses.A CNAME record says that a name is an alias for another name. The provider controls what the target resolves to, which is why you never point a status page at a hard-coded IP address.
In a BIND-style zone file:
; zone example.com
status 300 IN CNAME acme-co.status-provider.example.
Note the trailing dot after the target. Without it, BIND appends the zone name and you get acme-co.status-provider.example.example.com, which does not exist. Web dashboards usually add the dot for you, but zone files imported from scripts or Terraform often do not.
Then verify from the outside:
dig +short status.example.com CNAME
dig +short status.example.com @1.1.1.1
dig +short status.example.com @8.8.8.8
dig +trace status.example.com
The first command should print the provider target, the next two should show the final addresses, and +trace shows the full path from the root servers to your nameservers, which is the quickest way to spot a stale delegation or a nameserver that still serves an old zone.
Not with a plain CNAME. The zone apex already holds SOA and NS records, and a CNAME cannot coexist with other records at the same name. If you chose a dedicated domain like examplestatus.com, either use a subdomain of it (status.examplestatus.com) or a DNS provider that supports ALIAS, ANAME or CNAME flattening at the apex.
Your provider cannot serve HTTPS for status.example.com until it holds a certificate for that name. Hosted platforms typically request one automatically through ACME once the CNAME points at them, proving control by answering an HTTP or TLS challenge on the hostname. Until that succeeds, visitors see a certificate warning or a default certificate for the provider's own domain.
Three things block issuance:
example.com. CAA 0 issue "some-ca.example" limits which authorities may issue for example.com and its subdomains. If your provider's CA is not listed, issuance fails silently from your point of view. Add an entry for that CA or set a CAA record on the status subdomain itself.Check the result:
curl -sI https://status.example.com | head -n 1
echo | openssl s_client -connect status.example.com:443 \
-servername status.example.com 2>/dev/null \
| openssl x509 -noout -subject -issuer -dates -ext subjectAltName
The subjectAltName must contain status.example.com. If it only lists the provider's domain, the certificate for your name has not been issued yet.
One more trap: if example.com sends an HSTS header with includeSubDomains, browsers that visited your main site will refuse plain HTTP and refuse to click through certificate warnings on the status subdomain. Finish the TLS setup before you announce the URL. Auto-renewal can still fail later, so put the status hostname on the same expiry checks as your product; the reasons are covered in why certificates still expire with auto-renewal.
A self-hosted status page on the same server, cluster or cloud account is a page that reports "all systems operational" until it disappears with everything else. Hosting it with a separate provider solves the obvious part. The less obvious part is the list of shared dependencies that remain after you move it.
Walk through this list and write down what your status page shares with your product:
Each shared item is a scenario in which you cannot reach or update the page. The admin login is the one teams forget: if posting an update requires your company SSO and the SSO is part of the outage, the page stays green. Keep at least one account with a separate login and two-factor method. For more on hosting and status levels, see these status page best practices.
If someone queried status.example.com before the record existed, resolvers cached the NXDOMAIN answer. How long depends on the SOA minimum value of your zone, which you can read with dig +short example.com SOA (the last number, in seconds). During that window some users see "site not found" while your own checks pass. Create the record before anyone tests the URL.
An old A record from a previous self-hosted page, or a TXT verification record placed at status.example.com, prevents the CNAME. Some dashboards reject the change, others accept it and serve inconsistent answers. Put verification TXT records at the name the provider specifies, which is usually a different label.
With a wildcard like *.example.com in place, a mistyped record name such as statsu.example.com still resolves, just to your main site. Everything looks fine in dig until you compare the answer to the expected target.
After a DNS migration, the old provider often still has a copy of the zone. Changes made there have no effect. Compare dig +short example.com NS with the provider you are editing.
An A record copied from the current target works until the provider moves infrastructure, then it breaks without warning. Always use the CNAME target.
DNS failures affect your product as much as your status page, so it is worth checking resolution continuously; the failure modes and check intervals are covered in DNS monitoring: which failures take sites down.
Hosted status pages handle the infrastructure separation for you, as long as DNS is also kept independent. In m0nitor, public status pages support a custom domain via CNAME, password protection for private pages, email subscribers, multiple languages and a public incident history. The status page documentation walks through creating the page, adding components, connecting the custom domain and managing subscribers, including the CNAME target to use. The DNS checks and mistakes above apply regardless of which provider you choose.
The CNAME is usually visible on public resolvers within minutes if the name was never queried before. Certificate issuance follows once the provider can validate the hostname. Delays of hours usually point to negative caching, a CAA restriction or a proxied DNS record rather than slow propagation.
The provider has not issued a certificate for your hostname yet, so it serves its default certificate. Check that the CNAME resolves correctly from public resolvers, that the record is DNS-only rather than proxied, and that no CAA record on your domain excludes the provider's certificate authority.
A subdomain is more recognizable, a separate domain at a different registrar and DNS provider is more resilient. If you use a subdomain, reduce the shared risk with secondary DNS for the main zone and a longer TTL on the status record.
Usually each status page has its own target or its own hostname setting at the provider, because the provider needs to know which page to serve and which certificate to present. Follow the target shown for each page instead of reusing one.