The Security+ PKI Chain of Trust, Link by Link
Most Security+ certificate questions tend to revolve around the same basic idea: pinpointing where the chain breaks. Once the chain can be traced from memory, the PKI portion of objective 1.4 shrinks a lot. Thr
Most Security+ certificate questions tend to revolve around the same basic idea: pinpointing where the chain breaks. Once the chain can be traced from memory, the PKI portion of objective 1.4 shrinks a lot.
Three certificates, two signatures
A website's certificate isn't trusted by itself. It was signed by an intermediate CA, and that intermediate was signed by a root CA. The browser or OS has the root preloaded in its trust store, and that root is the foundation of trust, the starting point for every verification.
Upon connecting, the client reviews the chain. It ensures the server certificate was signed by the intermediate, and that the intermediate was signed by a root the client already has. It also checks for any expiration or revocation along the way. If all steps are valid, the connection is trusted. If any step fails, the user sees a warning page.
That means there are three certificates with two signatures between them. Keep that in mind, and most distractors disappear on their own.
Why the root stays offline
People wonder why the intermediate is needed when the root could simply sign server certificates itself.
Losing the root is expensive. Each client relying on it would need a new one, hence the root CA stays offline, used just for signing intermediates. When an intermediate is compromised, the CA revokes it and issues a new one, but the root (and all that relies on it) stays steady. On the exam, "offline root" and "revocable intermediate" are the reasons behind the hierarchy, and a question may ask exactly what that design protects.
The missing intermediate
This is the scenario CompTIA tends to favor, as it confounds those familiar with the terminology but not the underlying concepts.
A server receives a new certificate. It is valid and not expired, and the clients recognize the root CA. Still, these same clients flag the certificate as untrusted.
The issue arises because the server is not transmitting the intermediate certificate. Clients typically come with built-in roots and generally do not store intermediates, so the server must provide its own certificate along with the intermediate that signed it. This allows the client to construct the chain leading to the root it trusts. Exclude the intermediate, and the chain has a gap in the middle. The fix lives on the server: install the full chain. There is no problem with the root, and the client does not need a copy of the server certificate in its own store.
An option suggesting the root has expired on the web server, or that the client lacks the server certificate, is the misleading choice.
No anchor at all
The opposite failure is a certificate lacking a trusted root. A self-signed certificate is signed by its own key, with no CA above it. It works well in a lab or when a company pushes it out to its own devices, but a public browser has no root to rely on. A third-party certificate signed by a public CA already in browser trust stores doesn't face that issue.
When a stem says the browser cannot find a trusted issuer for the signing authority, the piece that's missing is a valid root of trust in the client's trust store. A longer key or extra names on the certificate won't rescue a chain that ends nowhere.
Revocation is a link failing on purpose
A certificate might be invalid even if its dates are still valid if the certificate authority revoked it. Security+ lists three methods a client can use to check:
- A CRL (Certificate Revocation List) is a signed list of revoked serial numbers that the client downloads. It's straightforward, but it can become lengthy and outdated between updates.
- OCSP (Online Certificate Status Protocol) queries the CA in real time about one certificate. It's more current, but it adds an extra round trip and reveals which sites the user visits.
- OCSP stapling lets the server include a recent, CA-signed status in its handshake, so the client skips the extra query and there's no privacy leak.
A question that touches on privacy or handshake speed while mentioning revocation likely refers to stapling.
Where the CSR fits
The chain begins with a request. A CSR (certificate signing request) holds the subject details and the public key for the CA, while the matching private key remains on the server. The requester signs the CSR using that private key to confirm ownership. So when someone asks what goes to the CA, the public key within the CSR is what travels. The private key stays with the server.
Practice the walk
What makes this work is speaking the chain out loud on every certificate question, from the server through the intermediate to the root in the trust store, then verifying dates and checking revocation. Then ask which link the stem broke. Our free Security+ practice questions are a good way to practice that habit, the Security+ objectives page shows where 1.4 fits in the full list, and the glossary covers CRL, OCSP, and the rest when an acronym slips.
Originally published on secplusmastery.com.
Originally published by Dev.to Security. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.