There seems to be some recurring confusion among Gentoo developers regarding the topic of OpenPGP key expiration dates. Some developers seem to believe them to be some kind of security measure — and start arguing about its weaknesses. Furthermore, some people seem to think of it as rotation mechanism, and believe that they are expected to generate new keys. The truth is, expiration date is neither of those.
The key expiration date can be updated at any time (both lengthened or shortened), including past the previous expiration date. This is a feature, not a bug. In fact, you are expected to update your expiration dates periodically. You certainly should not rotate your primary key unless really necessary, as switching to a new key usually involves a lot of hassle.
If an attacker manages to compromise your primary key, he can easily update the expiration date as well (even if it expires first). Therefore, expiration date does not really provide any added protection here. Revocation is the only way of dealing with compromised keys.
Expiration dates really serve two purposes: naturally eliminating unused keys, and enforcing periodical checks on the primary key. By requiring the developers to periodically update their expiration dates, we also implicitly force them to check whether their primary secret key (which we recommend storing offline, in a secure place) is still present and working. Now, if it turns out that the developer can’t neither update the expiration date nor revoke the key (because the key, its backups and the revocation certificate are all lost, damaged or the developer goes MIA), the key will eventually expire and stop being a ‘ghost’.
Even then, developers argue that we have LDAP and retirement procedures to deal with that. However, OpenPGP keys go beyond Gentoo and beyond Gentoo Infrastructure. We want to encourage good practices that will also affect our users and other people with whom developers are communicating, and who have no reason to know about internal Gentoo key management.
One thought on “OpenPGP key expiration is not a security measure”