Check DANE

The email domain. Its MX hosts are found automatically; DANE for SMTP is evaluated per MX host name (_25._tcp.<mx-host>), not for the email domain itself.

Advanced: specific SMTP host, port or address

With a host name, MX records are not used and the TLSA record is looked up for that host. An IP address tests one server directly (public addresses only).

TLSA Record Generator

Calculates TLSA association data from a certificate — locally in your browser. The result is a generated record, not one published in DNS: NetTools never changes DNS records.

How it works

NetTools finds the MX hosts, validates the DNSSEC chain of trust from the root itself for the MX, address and TLSA records, connects to each mail server on port 25 with STARTTLS in an isolated worker, captures the certificate chain the server presents, and evaluates every TLSA record with the semantics of its usage: DANE-EE matches the server key, DANE-TA verifies the chain to the published trust anchor and checks the name.

DANE and traditional CA validation

  • Traditional (PKIX) validation trusts a certificate because a certificate authority in the client's trust store signed it, and the name matches.
  • DANE publishes certificate associations as TLSA records in DNS, and DNSSEC authenticates those records. Without a valid DNSSEC chain a TLSA record proves nothing.
  • Usages 0 (PKIX-TA) and 1 (PKIX-EE) keep the normal CA requirements and add a constraint. Usages 2 (DANE-TA) and 3 (DANE-EE) make the TLSA record the trust anchor: DANE-TA names the issuing CA (which may be private), DANE-EE names the server's own certificate or key, without a public CA chain, name check or expiry check.
  • For SMTP between mail servers (RFC 7672), usages 0 and 1 may be treated as unusable; publish 3 1 1 (recommended) or 2 0 1. DANE does not switch off TLS itself: STARTTLS remains mandatory and protocol checks still apply.

Changing certificates safely (rollover)

  • Publish the TLSA record for the new key or certificate before deploying it, alongside the current one (RFC 7672 §4).
  • Wait at least twice the TLSA record's TTL so caches see both records; keep DNSSEC signatures valid throughout.
  • Then install the new certificate, verify with this checker, and only afterwards remove the old record.
  • 3 1 1 records survive renewals that keep the same key — reusing the key avoids most TLSA changes.

Safeguards & responsible use

  • Connections go only to public addresses of the MX hosts on SMTP ports (25, 465, 587, 2525). Each session sends only EHLO, STARTTLS and QUIT — never MAIL, RCPT, DATA or AUTH — and runs in an isolated worker thread with time and memory limits.
  • At most 5 MX hosts and 4 addresses per host, 4 concurrent probes server-wide, 10 probes per mail server address per 10 minutes, one running check per visitor. Results are kept in memory for 15 minutes.

Frequently asked questions

Does a TLSA record mean DANE works?

No. The record must be DNSSEC-authenticated, the server must offer STARTTLS, and the certificate it presents must satisfy the record. The checker tests all three.

Which TLSA record should I publish for SMTP?

RFC 7672 recommends “3 1 1” (DANE-EE, SubjectPublicKeyInfo, SHA-256), with “2 0 1” (DANE-TA, certificate, SHA-256) as the second choice.

Why are usages 0 and 1 marked unusable?

RFC 7672 says SMTP clients may treat PKIX-TA(0) and PKIX-EE(1) as unusable. The checker still evaluates them, but they do not count for SMTP DANE.