FHRP N10-009

Slide 1 of 8  |  N10-009 Obj 2.4  |  High Availability
FHRP — First Hop
Redundancy Protocols
One virtual gateway. Multiple physical routers. Automatic failover.
Your default gateway router fails. 500 users lose internet access because their gateway IP is gone. FHRP gives you a virtual IP that survives router failure — automatic failover with no client reconfiguration.
8 Slides N10-009 Obj 2.4 HSRP / VRRP / GLBP High Availability
Slide 2 of 8
The Single Point of Failure Problem
Every client hard-codes one default gateway. That is the problem.
Without FHRP
Every device on the subnet has one gateway configured: 10.0.0.1. That IP belongs to Router A — one physical interface on one device. Router A fails at 2 AM. 10.0.0.1 is gone. All 500 clients stop routing. No internet, no inter-VLAN, nothing. They will not recover until someone manually changes 500 devices or Router A comes back.
With FHRP
Clients configure the virtual IP: 10.0.0.1. That IP is shared between Router A (active) and Router B (standby). It does not belong to either physical interface. Router A fails. Router B detects the failure within seconds and assumes the virtual IP. Clients never know. No reconfiguration. Seamless.
Virtual Gateway
10.0.0.1
|
Router A (Active)
10.0.0.2
Router B (Standby)
10.0.0.3
|
500 Clients
gateway: 10.0.0.1
PC1 PC2 PC3 VIP 10.0.0.1 Router A Active Router B Standby Internet ISP
Clients always target the virtual IP.
Physical router IPs are irrelevant to them.
Slide 3 of 8
HSRP — Hot Standby Router Protocol
Cisco proprietary. Active/standby model. The most common exam protocol.
How HSRP Works
One router is elected Active — it owns the virtual IP and virtual MAC, forwards all traffic.
One router is Standby — listens for hellos from the Active, ready to take over.

Hello timer: default 3 seconds — how often Active sends hellos
Hold timer: default 10 seconds — how long Standby waits before declaring Active dead
Failover happens within hold timer seconds of Active failure.
Priority + Preemption
Priority: 0–255. Higher wins the Active election. Default is 100.
Set Router A to priority 110 to ensure it stays Active.
Preemption: must be explicitly enabled. Without it, if Router A fails and Router B takes over, Router A will NOT reclaim Active when it returns — even with higher priority. With preemption enabled, Router A reclaims immediately.
HSRP Virtual MAC
HSRP uses a well-known virtual MAC: 0000.0C07.ACxx where xx is the HSRP group number in hex. Group 1 = 0000.0C07.AC01. Clients ARP for the virtual IP and get this virtual MAC. When failover occurs, the new Active responds to ARPs with the same virtual MAC — no ARP table update required on clients.
HSRP Versions
HSRPv1: 255 groups, multicast 224.0.0.2, UDP 1985
HSRPv2: 4096 groups, multicast 224.0.0.102, improved millisecond timers
Versions are not interoperable on the same group — both routers must match.
Exam tip: HSRP is Cisco proprietary. If the question specifies "open standard," the answer is VRRP. If the question asks about load balancing across gateways, the answer is GLBP.
Slide 4 of 8
VRRP — Virtual Router Redundancy Protocol
Open standard (RFC 5798). Works on any vendor. Master/backup model.
VRRP vs HSRP Differences
VRRP uses Master (not Active) and Backup (not Standby) — same concept, different names.

Key difference: the Master router's real IP can be the virtual IP. No separate virtual IP is required if you use the Master's physical IP as the VIP. HSRP always requires a distinct virtual IP.

Advertisement interval: default 1 second (faster than HSRP's 3s)
Master down interval: 3 x advert interval + skew time
Multicast: 224.0.0.18 | Protocol: IP protocol 112
Virtual MAC
VRRP virtual MAC: 00-00-5E-00-01-xx where xx is the VRRP group (VRID) in hex. Same concept as HSRP — clients ARP for the VIP and get the virtual MAC regardless of which physical router currently holds Master role.
When to Choose VRRP
Multi-vendor environment — Cisco ASA + Juniper SRX + Palo Alto. HSRP only runs on Cisco. VRRP runs on everything. RFC 5798 (VRRPv3) also supports IPv6 natively. Any new deployment not locked into Cisco should default to VRRP.
Priority + Preemption
Priority: 1–254. Default 100. Higher wins Master election.
Priority 255 is reserved for the router that owns the virtual IP address physically.
Preemption is enabled by default in VRRP — opposite of HSRP where you must explicitly enable it. A returning high-priority router will automatically reclaim Master.
VRRP preemption is ON by default. HSRP preemption is OFF by default. This difference appears on CompTIA exams. Know it.
Slide 5 of 8
GLBP — Gateway Load Balancing Protocol
Cisco proprietary. The only FHRP that actively distributes traffic across multiple gateways.
How GLBP Differs
HSRP and VRRP: one router forwards, the rest idle. Standby routers sit unused.

GLBP: all routers forward simultaneously. One virtual IP is shared but each router gets its own virtual MAC. When a client ARPs for the virtual IP, the AVG (Active Virtual Gateway) assigns a different virtual MAC to each client using round-robin or weighted load balancing. Client A uses Router A's MAC. Client B uses Router B's MAC. Both gateways carry traffic simultaneously.
GLBP Roles
AVG (Active Virtual Gateway): owns the virtual IP, responds to ARPs, assigns virtual MACs to other routers. One per group.

AVF (Active Virtual Forwarder): each router in the group. Each AVF has its own virtual MAC and forwards traffic for the clients assigned to it. Up to 4 AVFs per GLBP group.
Load Balancing Methods
Round-robin: assigns virtual MACs in rotation. Equal distribution.
Weighted: assigns more MACs to higher-weighted routers. Proportional distribution.
Host-dependent: same client always gets the same virtual MAC. Useful for stateful applications.
When to Choose GLBP
Multiple gateway routers with spare capacity that is currently idle. High-traffic environments where one active gateway is a throughput bottleneck. Cisco-only environment — GLBP is proprietary, no open-standard equivalent.
GLBP virtual MAC format: 0007.B4xx.xxyy where xx.xx is the group number and yy is the AVF number. The AVG assigns each AVF's MAC to clients via ARP replies.
Inter- net VIP: 10.0.0.1 (shared) PC1 PC2 PC3 Router A AVF1+AVG Router B AVF2 Router C AVF3 All 3 routers active simultaneously
Slide 6 of 8
Virtual IP + MAC — How It Actually Works
Clients talk to the virtual MAC, not the router's real MAC.
Normal ARP Process (No FHRP)
Client needs to send traffic to internet. Destination is outside subnet — use gateway.
Client ARPs: "Who has 10.0.0.1?" Router A replies: "I do. My MAC is AA:BB:CC:DD:EE:01."
Client caches that mapping. Traffic flows via Router A's real MAC. Router A fails. ARP entry expires. Client re-ARPs. No reply. Connection dies.
FHRP ARP Process
Client ARPs: "Who has 10.0.0.1?" Active/Master router replies with the virtual MAC — not its real MAC.
Client caches: 10.0.0.1 = virtual MAC (e.g., 0000.0C07.AC01 for HSRP group 1).
Router A fails. Router B takes over and now responds to ARPs using the same virtual MAC. Client's ARP cache still valid. Traffic flows without interruption. Client never knows.
Gratuitous ARP on Failover
When the new Active takes over, it sends a gratuitous ARP — an unsolicited ARP reply advertising the virtual IP/MAC mapping. This forces immediate update of ARP caches on all devices in the subnet, including switches (which update their MAC address tables to point to the new physical port). No waiting for ARP entries to expire.
Why Separate Virtual MACs Matter
Switches maintain MAC address tables mapping MAC addresses to physical ports. Without virtual MACs, the switch would need to relearn which port the real MAC is on after failover. With virtual MACs, the gratuitous ARP updates the switch table to point the virtual MAC to the new physical port. Failover is sub-second for frames already on the wire.
Slide 7 of 8
Failover Process + Configuration Basics
What happens second-by-second when the active gateway fails.
T=0 Router A fails Power loss, link down, or hardware fault — hello packets stop transmitting
T=0 to 10s Hold timer counts down Router B expects a hello every 3s. After 10s of silence, it declares Active dead
T=10s Router B transitions to Active Assumes virtual IP and virtual MAC. Sends gratuitous ARP to update all caches
T=10s+ Traffic resumes Clients traffic now flows through Router B. TCP sessions that were mid-flight are dropped — must reconnect
Recovery Router A returns Without preemption: Router B stays Active. With preemption enabled: Router A reclaims Active (higher priority wins)
Clients → VIP Router A FAILED Router B Standby Internet Gratuitous ARP broadcast Normal operation — Router A forwarding
HSRP Config Example (Cisco IOS)
interface GigabitEthernet0/0
 ip address 10.0.0.2 255.255.255.0
 standby 1 ip 10.0.0.1      ! VIP
 standby 1 priority 110   ! Win election
 standby 1 preempt       ! Reclaim on return
 standby 1 timers 3 10   ! hello hold
Timer Tuning
Default hold timer is 10 seconds — that is 10 seconds of gateway outage. Tuning to sub-second timers (1ms hello, 3ms hold) dramatically reduces failover time but increases CPU and bandwidth overhead on the routers. Most production deployments use 1s/3s as a balance.
Slide 8 of 8  |  N10-009 Obj 2.4
HSRP vs VRRP vs GLBP: Exam Comparison
Feature HSRP VRRP GLBP
Standard Cisco proprietary RFC 5798 (open) Cisco proprietary
Active role name Active Master AVG / AVF
Standby role name Standby Backup Backup AVG
Load balancing No (one active) No (one master) Yes — all routers forward
Virtual MAC 0000.0C07.ACxx 00-00-5E-00-01-xx 0007.B4xx.xxyy
Multicast address 224.0.0.2 (v1) / 224.0.0.102 (v2) 224.0.0.18 224.0.0.102
Preemption default Off (must enable) On (enabled by default) Off (must enable)
Priority range 0 – 255 (default 100) 1 – 254 (default 100) 1 – 255 (default 100)
IPv6 support HSRPv2 only VRRPv3 native Limited
1 FHRP solves the single-point-of-failure gateway problem using a virtual IP shared between physical routers.
2 HSRP = Cisco, Active/Standby. VRRP = open standard, Master/Backup, preempt ON by default. GLBP = load balances across all routers.
3 Failover: Active fails → hold timer expires → Standby claims virtual IP → gratuitous ARP updates all caches. TCP sessions drop and must reconnect.