Design and build a privately hosted Public Key Infrastructure

The root CA is at the top of the hierarchy, this makes it a very attractive target for potential attackers.

If the root CA were to be compromised, an attacker could gain control of the entire PKI and compromise trust in the entire system, including any sub-systems reliant on the PKI.

It is considerably more difficult to replace a root CA compared to an Intermediate CA. If you have to replace a Root CA because of compromise this will impact the whole PKI, as you would need to replace the root CA in every end entities trust store. Replacing the root CA in the trust store of every end entity can be a time consuming as it could lead to support staff having to physically access every end entity.

Given its importance and potential disruption if compromised, the root CA must be kept offline (not connected to a network or disabled), should be unavailable for use and not allowing the issue of new certificates. It should only be used when it is required to complete an authorised function which should be audited and witnessed by a number of people.

You could, for example, store the root key on a secure encrypted token, or number of tokens. This key could then be used for an authorised function and require a number of people to complete the task. Keeping the root CA offline will provide separation between the root CA and the rest of the PKI, limiting its exposure.

In the event of an intermediate CA being compromised, you can bring the root online to issue a new certificate and revoke all certificates issued by the compromised CA. The intermediate CA certificate will also need to be revoked.