What is the difference between fail-open and fail-closed?
A fail-open device lets traffic through when it fails, favoring availability. A fail-closed device blocks traffic when it fails, favoring security.

Exam objectiveGiven a scenario, apply security principles to secure enterprise infrastructure
Objective 3.2 covers how to apply security principles to enterprise infrastructure: where security devices go, how they fail, which appliances filter and inspect traffic, how devices earn access to a port, and how remote users and sites connect safely.
Design decisions. Device placement, security zones, and attack surface decide how much an attacker can reach. A fail-open device favors availability and a fail-closed device favors security. An active device responds and a passive one watches; an inline device can reliably drop packets but can also break the path, while a tap or monitor port sees only copies.
Appliances and firewalls. A jump server controls admin access, a proxy server mediates requests, an IDS detects and alerts while an IPS adds prevention, a load balancer spreads requests, and sensors collect data. A web application firewall (WAF) inspects HTTP and HTTPS requests and blocks application-layer attacks such as SQL injection and cross-site scripting, a unified threat management (UTM) appliance bundles many functions, and a next-generation firewall (NGFW) adds application-aware deep packet inspection, often with user identity and intrusion prevention. Layer 4 filtering decides on address, port, and protocol, while Layer 7 filtering understands the application protocol.
Access and connectivity. Port security with 802.1X makes a device prove its identity before a switch port or access point admits it, with EAP carrying the authentication method. VPNs use IPsec or TLS tunnels for remote access and site links. SD-WAN connects sites over cheap links, and SASE adds cloud-delivered security near each user.
Exam scenarios give a requirement, such as blocking attacks inline, authenticating before granting a switch port, or securing branch offices, and ask for the right device or technology.
This page is the map. Knotch is where you drill it until it sticks.
An IDS primarily detects and alerts. An IPS adds prevention. Inline deployment is what lets an IPS reliably drop packets, while an out-of-band sensor can only respond indirectly. Incorrect prevention decisions can interrupt legitimate traffic.
An inline device has traffic pass through it, so it can reliably drop packets but is also a failure point. A tap or mirror-port device receives copies, so it cannot cause an outage and cannot directly drop packets.
SD-WAN is the overlay that links sites under central control. SASE layers cloud-delivered security on top, enforced near each user.
IPsec works at the network layer and protects IP packets between endpoints. TLS secures one application's session over TCP, usually on port 443, which most firewalls already allow, so TLS VPNs pass through more easily.
The full set of look-alike pairs for this objective is inside Knotch.
A team installed an intrusion prevention system on a switch mirror port. It raises alerts about SQL injection attempts, but the attacks still reach the web server.
Which change lets the device block these attacks?
Only an inline device can reliably drop packets before they reach the server.
An inline IPS can directly drop matching packets before they reach the server. A sensor on a tap or mirror port receives copies and cannot discard the original packets itself. At most it can send TCP resets or ask another device to block traffic, and those responses come after the first packets have already arrived, so they are not reliable enforcement. The trade-off is that an inline device can also become a failure point.
Why not B: Better signatures improve detection, but a device on a mirror port still sees only copies.
Why not C: Another mirror port still delivers copies, so the device cannot stop the original packets.
Why not D: An IDS only alerts, so it would block even less than the current device.
A fail-open device lets traffic through when it fails, favoring availability. A fail-closed device blocks traffic when it fails, favoring security.
Their position in the path tells them apart. A forward proxy represents clients, and a reverse proxy represents servers.
The test is depth versus breadth. An NGFW emphasizes application-aware deep inspection, often tied to user identity, while a UTM bundles many functions in one appliance, often for a smaller site that wants simplicity.
