NSE 8 flashcards
NSE 8 · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Trust boundary
- A line across which the required security policy changes. Zones are defined by trust boundary rather than physical location, so an exposed internal host belongs with other exposed assets.
- Blast radius
- The set of hosts reachable from a compromised segment. It equals the reachable set defined by policy, so many small zones with explicit allows shrink it.
- Asymmetric routing
- A flow whose forward and return packets traverse different firewalls. Stateful inspection fails because each device sees only one direction; fix by making the path symmetric or adding an exception.
- Session capacity sizing
- Estimate concurrent sessions as peak users times average flows per user, plus server connections, then add headroom for failover so a surviving HA member can absorb full load.
- TLS forward proxy
- Outbound decryption where the firewall presents an enterprise-CA certificate to clients and re-encrypts to servers. Requires clients to trust the enterprise CA.
- Inbound TLS decryption
- Decrypting server-side traffic requires the private key on the appliance. Certificate pinning on clients breaks this, so it is often limited to specific server pools.
- Hub-and-spoke VPN
- A topology where all sites tunnel to a central hub. Policy is centralized but hub bandwidth and one extra hop become bottlenecks as sites grow.
- Full mesh VPN
- Every site tunnels directly to every other site, giving n(n−1)/2 tunnels. Lowest latency but configuration and key management scale poorly.
- Out-of-band management
- A dedicated management segment reachable only from jump hosts, with no inbound path from user zones. Protects device administration even when a user zone is compromised.
- Split-brain HA
- Both HA members believe they are active because the heartbeat link failed. The design must fail closed, never open, when this occurs.
- Policy evaluation order
- Firewalls evaluate rules top-down and stop at the first match. Specific deny rules must sit above broad allow rules, with the implicit deny last.
- NAT evaluation order
- Destination NAT is applied before routing and before source NAT; source NAT is applied after the egress interface is chosen by the route lookup.
- Virtual IP requirement
- Inbound access via a VIP needs both the VIP object (address and port mapping) and a firewall policy permitting the translated traffic. One without the other silently fails.
- Route precedence
- Policy routes override static routes; static routes are chosen by administrative distance then metric; connected routes win over all dynamic and static entries.
- Policy route pitfall
- A policy route silently overrides a correct static route. When traffic exits an unexpected interface, check policy routes before examining the routing table.
- Interface-mode IPsec
- The tunnel is bound to a virtual interface and ordinary firewall policies control traffic through it. Preferred over policy-mode for complex routing and multiple subnets.
- Policy-mode IPsec
- The tunnel is bound directly to a firewall policy. Mixing policy mode and interface mode for the same peer causes phase 2 selector mismatches and failed tunnels.
- Phase 1 parameters
- IKE version, encryption, hash, DH group, pre-shared key, and peer ID must match on both ends. Any mismatch prevents the management tunnel from forming.
- Phase 2 selectors
- The local and remote subnet pairs negotiated for data. A phase 2 failure with phase 1 up almost always indicates mismatched or overlapping selectors.
- Tunnel MTU clamping
- IPsec adds overhead, so tunnel interfaces use reduced MTU, typically 1400 or lower, and TCP MSS is clamped to prevent fragmentation of large packets and routing hellos.
- VDOM isolation
- Virtual domains separate routing tables and policies per tenant or function. Traffic between VDOMs requires an inter-VDOM link plus an explicit firewall policy.
- SD-WAN SLA probe
- Probes measure latency, jitter, and loss per member link to steer applications. If no member meets the SLA, traffic follows the defined fallback, which must be explicit.
- Config diff before change
- Export and diff the running configuration against the last known-good backup before and after changes, so a regression can be identified quickly.
- Layered isolation
- Troubleshoot in order: link and ARP, then route selection, then policy match, then session and NAT state. Stop at the first failing layer.
- No session created
- If no session exists, the packet never matched a policy or the route lookup failed. Check policy match keys and the return route before deeper packet analysis.
- Session without reply
- A session exists but shows no return traffic. Causes include wrong NAT translation, a server default gateway pointing elsewhere, or an intermediate firewall on the return path.
- Phase 1 up phase 2 down
- The IKE tunnel formed but data selectors failed. Check for mismatched or overlapping local and remote subnets, or a duplicate phase 2 entry.
- MTU symptom pattern
- Small pings succeed while large transfers stall. Test with DF-set pings of decreasing size to find path MTU, then clamp TCP MSS on the tunnel.
- Inbound works outbound fails
- A classic NAT or gateway error: the server's default gateway does not point at the firewall, or source NAT is applied in only one direction.
- Certificate chain validation
- Clients must trust the root and receive the intermediate certificates. A missing intermediate causes failures that look like connectivity problems in logs.
- Log correlation
- Compare firewall traffic logs, event logs, and server logs by timestamp and source port. An allow followed by a reset means the server refused, not the firewall.
- HA sync mismatch
- One member behaves differently from its peer after a change. Check HA configuration synchronization status before diagnosing packet-level causes.