Information Security Stack Exchange is a question and answer site for information security professionals. It only takes a minute to sign up.
Sign up to join this community
I am trying to understand meaning of the Subject Alternative Name extension. The reason of usage it in SSL X.509 certificates or other end-entities certificates is mostly clear for me.
However, I didn't find any information whether it is feasible or not to use this extension in CAs certificates (self-signed or not). Is it possible in general? What can be the reason if yes?
The subject name, in a certificate, is in the subjectDN field, which is a Distinguished Name: a structure name whose initial goal was to serve as index in the Directory hierarchy (the Directory never existed in practice, but it can be imagined as a giant worldwide LDAP server).
Sometimes, there are some usages which call for notions of identity that do not map well (or at all) to the format of a DN. For instance, if a certificate is to be used for S/MIME, then it is desirable to encode email addresses in certificates, and the standard way to do that is through a Subject Alt Name extension.
As a rule, what matters is what applications will do. You can put a Subject Alt Name extension with arbitrary contents in a CA certificate. However, it is mostly useless since CA certificates are meant to be processed by applications who validate certificate chains -- these applications will ignore the SAN extension in a CA certificate. To put things briefly: the notion of identity used by CA certificates is the one needed for building and validating certificate chains, and that uses the DN (from the subjectDN and issuerDN fields). Thus, SAN extensions are not useful for CA certificates.
(Well, mostly. Theoretically, you could use Name Constraints extensions that apply restrictions on subsequent certificates in the chain, including CA certificates, and these constraints take the SAN into consideration. However, Name Constraints, though standard, suffer from lack of support from usual libraries, so in practice nobody uses them.)
(Still theoretically, it is allowed by standards to have an empty subjectDN, in which case a SAN extension must be present and contain a name of type directoryName, i.e. a DN, and that DN will then be used as if it had been found in the subjectDN field. The purpose of this specific point is unclear; I guess it was included to appease the aesthetic sense of some committee member, somewhere.)
Maybe I am misunderstanding the question, but one of the most common reasons one would use a SAN cert is for sites that have a CNAME or alias.
For example:
If a user accesses https://example.com but the certificate is issued to www.example.com then the user will receive a name mismatch warning.
If a SAN cert is used then both domains can be specified as valid and the user will not receive a warning.
