The idea in one sentence
Your mail server adds a cryptographic signature to every outgoing message. The recipient fetches your public key from DNS and checks whether the signature matches.
What exactly is signed?
DKIM forms a hash over a configurable part of the message — typically the body and important headers like From, To, Subject, Date. The hash is signed with your private key and attached to the mail as an extra DKIM-Signature header.
Example DNS record
; TXT record for selector1._domainkey.firma.at
"v=DKIM1; k=rsa; p=MIGfMA0GCSqGSIb3DQEBAQUAA4GN…"
The key fields:
v=DKIM1— DKIM version.k=rsa— algorithm (usually RSA, increasingly also Ed25519).p=…— the public key, often 1024 or 2048 bits.
Selectors — why?
A selector is a name placed before ._domainkey. in DNS. It lets you operate multiple keys at the same time — handy for key rotation or when several services (e.g. CRM and mail server) sign separately.
DKIM alone is not enough
DKIM proves: this message was issued by someone with access to the signing domain's key and was not altered in transit. But it says nothing about whether the visible From address matches the signing domain. Only DMARC checks this alignment.
Practical tips
- Key length ≥ 2048 bits — 1024 is considered outdated.
- Key rotation every 6–12 months — using separate selectors, without downtime.
- Each sending service needs its own selector — newsletter tool, CRM, mail server.