How Do You Audit Your DNS for Dangling Records and Orphaned Subdomains?
Published 2026-11-14 · Reviewed 2026-11-14 · 6 min read · By the Ingenious AI team
Is someone impersonating your domain right now?
Run the free DomainGuard scan: lookalike registrations, MX-armed variants, certificate-transparency and CAA checks for your domain, in one report.
Run my free DomainGuard scanThe short answer
You audit DNS by listing everything you point at and everything that points at you: export the full zone, classify every record that references an external service, resolve each chain to what actually answers today, and cross-check your subdomains against the certificates issued for them. Every record then gets a verdict — owned and current, owned and stale, or orphaned. Orphaned records are deleted the day they are confirmed; stale ones get owners and dates. The audit takes a day for a typical small estate and removes the raw material of subdomain takeovers and shadow services.
Why DNS zones rot
A DNS zone is a diary nobody meant to keep. Every campaign microsite, vendor onboarding, verification token, developer experiment, and agency project added records; almost none were removed when the project ended, because removal has no owner and no deadline. Over years the zone accumulates two dangerous categories: records pointing at things that no longer exist (dangling), and subdomains still live but forgotten by everyone except the internet (orphaned). Both are inventoried continuously by attackers, whose scanners have more patience for your history than your own team does.
The audit's value is that it converts that history back into a managed present. Companies are consistently surprised by the first export — subdomains for events three rebrands ago, TXT records for tools nobody can name, a mail record for a service cancelled in a previous decade. None of it is shameful; all of it is attack surface with a delete key. The only shameful version is the second year in a row it stays unexamined after somebody printed the list.
Step one: the full inventory
Export the zone from every provider you use — registrar DNS, a dedicated DNS host, cloud DNS — because estates often sprawl across two or three. List every record with type, name, target, and TTL. Then sort by intent: apex and mail fundamentals (MX, SPF, DKIM, DMARC), service records for the live site and app, verification and vendor TXT records, and the long tail of CNAMEs and subdomains. The tail is the audit's real subject, so give every tail record a row in a working document with space for owner, purpose, and verdict.
Add the external view while you inventory. Enumerate your subdomains from Certificate Transparency logs and passive DNS — they will include names missing from your export, because they were created in a provider account someone forgot, or under a delegated subdomain managed by a vendor. Delegations (NS records handing a whole subdomain to another provider) deserve special suspicion: each one is a zone-within-the-zone that your export does not show and your audit must chase separately.
Step two: resolve everything to ground truth
For every external pointer, follow the chain to what answers today. A live service you recognise and can administer: current. A provider's unclaimed-resource error — the storage bucket that does not exist, the app hostname unassigned: dangling, and a takeover risk ranked by the subdomain's plausibility (an orphaned login or mail-shaped subdomain outranks an old campaign name). A response you do not recognise at all: investigate before anything else, because unknown live content under your name is an incident until explained, not an audit footnote.
Cross-check certificates as the second source of truth. CT logs list every certificate issued for your names; filter to issuances no team claims. A certificate for a subdomain absent from your zone export means a delegation or a provider account outside your map. A certificate from a CA you never use, on a name you do use, means someone satisfied domain validation through a channel you do not control — the dangling-record path is the usual explanation, and finding it in an audit rather than an incident is the entire point.
Step three: verdicts, deletions, and documentation
Work the list to verdicts in one pass. Dangling and confirmed-dead: delete immediately, in the audit session, not in a follow-up ticket — deletion is reversible in minutes if you misjudged, while exposure compounds daily. Live but ownerless: assign an owner on the spot or schedule the service's retirement; an orphan subdomain running an unpatched anything is a foothold with your brand on it. Current and owned: record the owner and the renewal or review date next to the record in your register, so the next audit starts from documentation instead of archaeology.
Handle mail records with extra care in both directions. Forgotten MX records on subdomains can receive real mail into dead systems — a privacy incident in slow motion — while deleting a live MX in error stops mail outright. Verify mail flows with the business owner before touching MX, and afterwards, publish the lockdown set (null MX, reject DMARC) on subdomains that should never send or receive, so their intent is explicit to the world.
Making the audit a rhythm, not an event
The first audit is archaeology; the value compounds when it becomes a calendar item with a shrinking scope. Quarterly, re-export and diff against the register: new records should all have owners (because creation now requires one), deletions should match decommissioning tickets, and any surprise is small enough to chase the same day. Tie the process changes to the audit's findings — the paired-ticket rule for resource deletion, the named-owner rule for new records — and the zone stops re-rotting.
Fold the results into your wider posture reporting. Dangling records found and removed, orphaned services adopted or retired, delegations mapped: these are concrete, countable improvements an insurer or enterprise customer can verify, and they belong next to your DMARC enforcement and monitoring numbers as evidence the estate is managed. DomainGuard's external scans see much of this surface from outside — takeover-risk findings in a report are the same list, compiled the way an attacker compiles it — which makes the report a useful cross-check that your internal audit missed nothing.
Five actions to take today
- Export every zone from every DNS provider and list all external pointers with owner and purpose columns.
- Enumerate subdomains from CT logs and passive DNS to catch delegated or forgotten names your export misses.
- Resolve every CNAME chain; flag unclaimed-resource errors as dangling and delete confirmed-dead records the same day.
- Cross-check certificates issued for your names and investigate any issuance no team can claim.
- Convert the audit into a quarterly diff against a register, with named-owner and paired-deletion rules for all future changes.
Worked example: Mesa Verde Rentals cleans six years of DNS in a day
Mesa Verde Rentals, a fictional property company monitoring 4 domains, exported zones totalling 486 records. The audit found 19 dangling CNAMEs — including pointers to a booking engine it had left two platforms ago — 3 orphaned subdomains still serving forgotten campaign pages, and 1 delegated subdomain at a vendor nobody had a contract with anymore. All 19 dangling records were deleted in the session, 2 orphans were retired, 1 was adopted and patched, and the delegation was reclaimed. The register that replaced the archaeology fits on two pages, and the next quarterly diff took 40 minutes.
FAQ
What tools do I need for a DNS audit?
Your providers' zone exports, a DNS lookup tool (dig or an online equivalent), a CT search such as crt.sh, and a spreadsheet. The method matters more than the tooling; everything required is free.
Is deleting a record risky?
For a confirmed-dead target, the record serves nobody legitimate — that is what dead means. Verify with the owning team, snapshot the zone export first so any record can be restored in minutes, then delete.
How do I find subdomains I have forgotten entirely?
Certificate Transparency logs and passive DNS services enumerate names that have ever had certificates or been observed. Compare that list against your zone export; the difference is your forgotten estate.
What is a delegated subdomain and why chase it?
An NS record delegating a subdomain hands its entire DNS to another provider or vendor. Your zone export shows only the delegation, not what lives beneath it — a classic hiding place for orphaned and dangling records.
How often should the full audit run?
A deep audit annually, with quarterly diffs against your register in between, and event-driven checks whenever a vendor, campaign, or platform is decommissioned.