Community hub
0 subscribers7 pages, 0 posts
Recent from talks
All channels
Be the first to start a discussion here.
Be the first to start a discussion here.
Be the first to start a discussion here.
Be the first to start a discussion here.
Contribute something
Welcome to the community hub built to collect knowledge and have discussions related to First-hop redundancy protocol.
Nothing was collected or created yet.
First-hop redundancy protocol
View on Wikipediafrom Wikipedia
First hop redundancy protocols (FHRP) are a category of networking protocols designed to protect the default gateway used on a subnetwork by allowing two or more routers to provide backup for that address.[1][2][3] In the event of failure of an active router, the backup router will take over the address, usually within a few seconds. In practice, such protocols can also be used to protect other services operating on a single IP address, not just routers.
Examples of such protocols include (in approximate order of creation):
- Hot Standby Router Protocol (HSRP) - Cisco's initial, proprietary standard developed in 1998[4]
- Virtual Router Redundancy Protocol (VRRP) - an open (albeit patent-encumbered) standard protocol based largely on Cisco's HSRP
- Common Address Redundancy Protocol (CARP) - patent-free, unencumbered alternative to Cisco's HSRP and the IETF's VRRP, developed in October 2003
- Extreme Standby Router Protocol (ESRP) - Extreme Networks' proprietary standard with fast failover and layer 2 protection
- Gateway Load Balancing Protocol (GLBP) - a more recent proprietary standard from Cisco that permits load balancing as well as redundancy
- Routed Split multi-link trunking (R-SMLT) - an Avaya redundancy protocol
- NetScreen Redundancy Protocol (NSRP) - a Juniper Networks proprietary router redundancy protocol providing load balancing
- Chassis Cluster Redundant Ethernet - a Juniper Networks proprietary Ethernet redundancy protocol, used on its SRX platform[5]
- Multi-active Gateway Protocol (MAGP) - a Mellanox proprietary protocol based on VRRP that allows active-active operation[6]
References
[edit]- ^ Wendell Odom; Jason Gooley; David Hucaby (5 August 2024). "First Hop Redundancy Protocols". Cisco Press.
The term First Hop Redundancy Protocol (FHRP) refers to the category of protocols that enable hosts to take advantage of redundant routers in a subnet.
- ^ "First Hop Redundancy Protocol (FHRP)". Cisco. Retrieved 2019-06-05.
- ^ gokhankosem (2018-10-16). "First Hop Redundancy Protocols". IpCisco. Retrieved 2019-06-05.
- ^ "Cisco Protocols: HSRP, VRRP, GLBP". Petri. 2012-04-11. Retrieved 2019-06-05.
- ^ "Chassis Cluster Redundant Ethernet Interfaces - TechLibrary - Juniper Networks". Juniper Networks. 2019-05-30. Retrieved 2019-06-09.
- ^ "HowTo Configure MAGP on Mellanox Switches". community.mellanox.com. 1 March 2020.
 | This computer networking article is a stub. You can help Wikipedia by expanding it. |
First-hop redundancy protocol
View on Grokipediafrom Grokipedia
| Protocol | Developer/Standard | Election Model | Load Balancing | Key Characteristics |
|---|---|---|---|---|
| HSRP | Cisco (proprietary) | Active/standby | Per-subnet (via multiple groups) | Uses priority-based election (default 100) for active router selection; routers exchange hello messages via multicast UDP to monitor status and trigger failover; supports virtual MAC addresses for seamless ARP responses.[2] |
| VRRP | IETF (RFC 5798, updated by RFC 9568) | Active/standby (Master/Backup) | None (focus on redundancy) | Open standard for IPv4 and IPv6; elects a Master router based on priority to forward packets for virtual addresses; Backup routers monitor via periodic advertisements and assume responsibility upon Master failure, enhancing availability without dynamic routing on end-hosts.[3] |
| GLBP | Cisco (proprietary) | Active/active | Per-host (up to 4 virtual forwarders) | Elects an Active Virtual Gateway (AVG) to assign virtual MAC addresses and distribute traffic; supports up to 1024 groups per interface with authentication options (MD5 or plain text); provides both redundancy and efficient load sharing across LAN devices.[4] |
Introduction
Definition and Scope
First-hop redundancy protocols (FHRPs) are a category of networking protocols designed to provide redundancy for the first-hop router, also known as the default gateway, in a subnetwork by enabling multiple routers to share a single virtual IP address (VIP), thereby ensuring seamless failover without host reconfiguration.[2] These protocols operate by electing one router as the active forwarder of traffic destined outside the local subnetwork, while others serve as standbys ready to assume responsibility upon failure. Key concepts include the VIP, which acts as the shared gateway address configured on hosts; a virtual MAC address used for Address Resolution Protocol (ARP) resolution to direct traffic to the active router; and group membership, where participating routers coordinate via multicast advertisements to maintain synchronization and detect failures.[5] Examples of FHRPs within this category include the Hot Standby Router Protocol (HSRP), Virtual Router Redundancy Protocol (VRRP), and Gateway Load Balancing Protocol (GLBP).[6] The scope of FHRPs encompasses both IPv4 and IPv6 environments, protecting against single points of failure in local area networks (LANs) or virtual LANs (VLANs) where end hosts depend on a default gateway for routing to external networks.[5] In IPv4 setups, protocols like VRRP version 2 (VRRPv2) and HSRP version 1 focus on shared IPv4 addresses, while extensions such as VRRP version 3 (VRRPv3), as updated by RFC 9568 in 2024, and HSRP version 2 explicitly support IPv6 address families, including link-local and global addresses, to maintain compatibility with modern dual-stack deployments.[2] This redundancy mechanism is particularly vital in multiaccess LAN environments, such as Ethernet, where multicast communications allow routers to form virtual router groups without requiring changes to host configurations. FHRPs emerged in the 1990s to address the growing need for gateway redundancy in enterprise networks as TCP/IP and router adoption expanded, with Cisco introducing HSRP in 1994 and the IETF standardizing VRRP in 1998.[7][8][9] This development marked a shift toward high-availability designs, mitigating downtime from default gateway failures in increasingly complex LAN infrastructures.[8]Purpose and Benefits
First-hop redundancy protocols (FHRPs) are designed to eliminate the single point of failure associated with default gateways in IP networks, thereby ensuring continuous connectivity for end hosts even during router outages or failures. By allowing multiple routers to share a virtual IP address as the default gateway, FHRPs enable automatic failover to a standby router without requiring manual intervention or reconfiguration of hosts. This redundancy mechanism is particularly vital in environments where the first-hop router serves as the critical path for all outbound traffic from local networks.[2] The primary benefits of FHRPs include rapid failover capabilities, with failover typically occurring in a few seconds; for example, HSRP defaults to about 10 seconds, while VRRP defaults to around 3 seconds, and both can be enhanced to sub-second levels when integrated with Bidirectional Forwarding Detection (BFD) for quicker failure detection. This results in significantly increased network uptime, approaching near-100% availability by minimizing disruptions and supporting high-availability clustering across redundant devices. Additionally, FHRPs simplify host configuration, as end devices need only point to the shared virtual IP address rather than individual router IPs, reducing administrative overhead and potential errors in large-scale deployments. These advantages collectively lower the mean time to repair (MTTR) for gateway failures from potential minutes of manual recovery to mere seconds of automated switching.[2][10][11] Beyond traditional routing, FHRPs extend protection to other IP-based services such as firewalls and load balancers by enabling shared virtual IPs for failover in clustered setups. For instance, protocols like HSRP can be used in firewall clusters to maintain consistent gateway addressing across sites. FHRPs also integrate with tracking features to monitor interface status or specific routes, triggering preemptive failovers for proactive redundancy. In open-standard environments, VRRP provides similar benefits for multi-vendor interoperability.[12][2]Core Protocols
Hot Standby Router Protocol (HSRP)
The Hot Standby Router Protocol (HSRP) is a Cisco-proprietary first-hop redundancy protocol developed in the mid-1990s to provide gateway redundancy for IP networks.[1] It enables multiple routers on a LAN to form a standby group that emulates a single virtual router, with one router acting as the active forwarder of traffic and others serving in standby roles.[13] HSRP groups communicate via multicast hello packets sent to the address 224.0.0.2 on UDP port 1985, allowing routers to monitor each other's availability and elect roles dynamically.[13] This architecture ensures seamless failover without host reconfiguration, though it supports only active-standby operation rather than load sharing across routers.[10] Key components of HSRP include a priority-based election mechanism, where the router with the highest priority (default 100, configurable range 0-255) becomes the active router, with ties resolved by the highest IP address.[13] Authentication is supported in plain-text form (default 8-character key "cisco") or enhanced MD5 for security against spoofing, where each group member generates a keyed hash included in hello packets.[13][14] HSRP exists in two versions: version 1 (HSRPv1) for IPv4 networks, which uses the multicast address 224.0.0.2 and virtual MAC addresses in the format 0000.0c07.acXX (where XX is the hexadecimal group number), and version 2 (HSRPv2) which extends support to IPv6 addresses while addressing limitations like expanded group numbering.[13][15] Configuration of HSRP on Cisco IOS devices involves enabling it on interfaces, assigning a virtual IP (VIP) address, specifying a group number (1-255 for v1, up to 4095 for v2), and optionally setting priority and timers.[16] For example, to configure group 1 with VIP 192.168.1.1 and priority 110 on an interface, the CLI commands are:interface GigabitEthernet0/0, standby version 1, standby 1 ip 192.168.1.1, standby 1 priority 110.[17] Preemption is enabled by default (via standby 1 preempt), allowing a higher-priority router to reclaim the active role upon recovery.[10]
Unique features of HSRP include object tracking, which integrates with Enhanced Object Tracking to monitor interface states, IP routes, or IP SLA probes; if a tracked object fails (e.g., a WAN link goes down), the router decrements its priority (default by 10) to trigger failover.[18] The virtual MAC address format ensures consistent Layer 2 identification for the group, facilitating ARP responses from the active router.[13] However, HSRP does not support load balancing across active routers and is interoperable only with Cisco devices due to its proprietary nature.[10] Like VRRP, HSRP relies on a similar priority-based election for role assignment.[10]
Virtual Router Redundancy Protocol (VRRP)
The Virtual Router Redundancy Protocol (VRRP) is an open-standard election protocol that provides redundancy for the default gateway in IPv4 and IPv6 networks by dynamically assigning responsibility for a virtual router to one of the participating routers on a local area network (LAN). Defined in IETF RFC 3768 in April 2004, which obsoletes the earlier RFC 2338 from March 1998, VRRP operates as an IP protocol with number 112 and uses multicast address 224.0.0.18 for advertisement messages sent from the master router to backup routers.[19] Developed as an open alternative inspired by Cisco's proprietary Hot Standby Router Protocol (HSRP), VRRP enables high availability without vendor lock-in.[19] Key components of VRRP include the election of a master router and one or more backup routers based on a configurable priority value, which ranges from 1 to 255 with a default of 100; the router with the highest priority becomes the master, and ties are broken by the highest IP address.[19] The master sends periodic advertisement messages at a default interval of 1 second to inform backups of its status, triggering failover if advertisements cease for a skew-adjusted period (typically 3 seconds).[19] VRRP version 2 (VRRPv2), specified in RFC 3768, supports IPv4 only, while version 3 (VRRPv3), defined in RFC 5798 (obsoleted by RFC 9568 in 2024), extends support to both IPv4 and IPv6 and allows multiple virtual routers per physical interface for enhanced flexibility.[3] Configuration of VRRP involves defining a virtual router ID (group number), a virtual IP address (VIP), and priority on participating routers' interfaces; for example, on Cisco or compatible multi-vendor devices, the commands might includeinterface vlan 1, vrrp 1 ip 192.168.1.1, and vrrp 1 priority 110 to set up group 1 with VIP 192.168.1.1 and elevated priority. Unique features distinguish VRRP from similar protocols: the "owner" concept designates the router whose real interface IP matches the VIP as having inherent priority 255, ensuring it defaults to master if operational; in "accept mode," the master can forward traffic destined to its own real IP addresses beyond just the VIP, unlike strict non-owner behavior that discards such packets.[19] Additionally, VRRP can integrate with Bidirectional Forwarding Detection (BFD) to accelerate failure detection beyond advertisement timeouts, enabling sub-second failover in implementations from vendors like Cisco and Juniper.[20]
As an IETF standard, VRRP offers significant advantages in interoperability across multi-vendor environments, including support from Cisco, Juniper, Nokia, and Huawei routers, without proprietary licensing restrictions that apply to alternatives like HSRP.[21][22][23] This openness promotes widespread adoption in enterprise and service provider networks seeking reliable first-hop redundancy.[19]
Gateway Load Balancing Protocol (GLBP)
Gateway Load Balancing Protocol (GLBP) is a Cisco proprietary first-hop redundancy protocol introduced in 2005 that enhances gateway redundancy by incorporating load balancing across multiple routers using a single virtual IP address and multiple virtual MAC addresses. It operates on IEEE 802.3 LANs, providing automatic failover while distributing traffic on a per-host basis to optimize bandwidth utilization. GLBP builds on redundancy concepts similar to HSRP but adds active-active forwarding capabilities through its core components: the Active Virtual Gateway (AVG), which handles ARP responses, and Active Virtual Forwarders (AVFs), which perform packet forwarding.[24][25] The AVG is elected among participating routers based on a configurable priority value (default 100, range 1-255), with the highest priority router winning; ties are broken by the highest IP address, and preemption can be enabled to allow higher-priority routers to take over. Up to four AVFs can be active per GLBP group, each associated with a unique virtual MAC address derived from the format 0007.b400.XXYY, where XX is the group number and YY is the forwarder number. Load balancing algorithms include host-dependent (assigning the same virtual MAC to a host for session consistency), round-robin (alternating assignments for even distribution), and weighted (proportional to router capacity via configurable weights). GLBP communicates via UDP port 3222, using multicast address 224.0.0.102 for hello messages sent every 3 seconds (default hold time 10 seconds).[25] Configuration of GLBP involves enabling the protocol on router interfaces within the same subnet, specifying the group number and virtual IP address. For example, the commandglbp 1 ip 192.168.1.1 sets up group 1 with the virtual IP, while glbp 1 priority 110 configures a higher priority for AVG election. Weights for AVFs are set with glbp 1 weighting 100 (default maximum), and lower/upper thresholds can limit participation based on tracked objects like interface status. GLBP supports multiple groups (up to 255) for further segmentation.[25]
A distinguishing feature of GLBP is its manipulation of ARP replies by the AVG to assign different virtual MAC addresses to individual hosts, enabling true per-host load distribution without requiring client changes. It integrates with Enhanced Object Tracking (EOT) to monitor interfaces, IP SLAs, or other objects, dynamically adjusting priorities or weights for failover triggers, similar to tracking in HSRP or VRRP. This setup ensures seamless redundancy while maximizing gateway efficiency. By allowing up to four AVFs to forward traffic concurrently, GLBP can improve bandwidth utilization up to fourfold compared to single-active protocols, reducing bottlenecks in high-traffic environments.[25]
Operational Mechanisms
Election Process and States
In first-hop redundancy protocols (FHRPs), the election process determines the active router responsible for forwarding traffic on behalf of a virtual router shared by multiple physical routers on the same LAN segment. Routers participating in an FHRP group exchange hello or advertisement packets via multicast IP addresses to advertise their presence, priority, and current state; these packets are sent at configurable intervals that vary by protocol—for example, every 3 seconds by default in HSRP and GLBP, or 1 second in VRRP—with hold times generally set to three times the advertisement interval (e.g., 10 seconds in HSRP, 3 seconds in VRRP), during which the absence of advertisements triggers state transitions.[26][27][28] The election favors the router with the highest priority value, ranging from 1 to 255, where a higher number indicates greater preference; in the event of a tie, the router with the highest IP address on the interface serves as the tiebreaker.[26][29] Routers form groups by joining the designated multicast group upon configuration of the virtual IP address, which represents the shared default gateway for end hosts; this virtual IP, along with a virtual MAC address, is advertised to hosts through gratuitous ARP packets sent by the active router to update local ARP caches without solicitation.[27][30] Common parameters include advertisement timers that synchronize group members on heartbeat intervals and hold-down timers that prevent premature state changes due to transient packet loss; additionally, priority can be dynamically incremented during preemption to allow a higher-priority router to assume the active role without waiting for a failure.[31] State management in FHRPs follows a finite state machine (FSM) to ensure orderly transitions and role assignments within the group, though the specific states vary by protocol. For example, HSRP includes states such as Initialize (or Init), where the router starts up and has not yet learned group details; Learn, during which the router gathers information from advertisements to understand the virtual router configuration; Listen, where non-electable routers monitor the group without participating in elections; Speak, a transitional state for routers advertising their intent to contend for active or standby roles; and Active and Standby, where the Active router forwards traffic using the virtual addresses while the Standby awaits failover. In contrast, VRRP uses a simpler model with Initialize, Master (equivalent to Active), and Backup (equivalent to Standby) states.[26][32][33] Transitions between states are triggered by events such as timeouts on hold timers, receipt of higher-priority advertisements, or interface status changes, with terminology varying slightly across protocols—for instance, HSRP uses Active/Standby while VRRP employs Master/Backup.[34][33]Failover and Preemption
In First Hop Redundancy Protocols (FHRPs), failover occurs when the active router fails to send hello messages within the configured hold timer, typically set to 10 seconds by default in protocols like Hot Standby Router Protocol (HSRP), prompting the standby router to promote itself to the active role.[35] Upon assuming the active role, the new active router transmits gratuitous Address Resolution Protocol (ARP) replies to update the ARP tables of connected hosts, ensuring traffic redirection to the virtual IP address without manual intervention.[10] Failure detection in FHRPs relies on periodic hello messages exchanged between routers, with the hold timer serving as the primary threshold for declaring a peer down.[36] Optional interface tracking enhances detection by monitoring specific interfaces or objects; if a tracked interface fails, the router's priority is decremented, potentially triggering an immediate failover without waiting for the hold timer.[10] For sub-second detection, FHRPs can integrate with IP Service Level Agreement (IP SLA) probes or Bidirectional Forwarding Detection (BFD), which provide rapid path failure notifications to accelerate role switches. Preemption is a configurable feature in FHRPs that enables a higher-priority router to reclaim the active role after recovering from a failure, overriding the current active router if its priority is lower.[36] To prevent network instability or flapping, preemption includes configurable delay timers, with a default of 0 seconds in HSRP and Virtual Router Redundancy Protocol (VRRP), allowing administrators to introduce brief stabilization periods post-recovery.[10] In VRRP, preemption is enabled by default per RFC 9568.[5] During recovery, the newly promoted active router begins sending periodic advertisement messages (hellos in HSRP or advertisements in VRRP) to inform backup routers of its status, causing them to adjust their states accordingly and maintain group synchronization. Without enhancements like BFD, typical failover times range from 1 to 5 seconds, depending on timer configurations, though this can extend to the full hold timer duration in default setups.[35] Edge cases in FHRP operations include split-brain scenarios, where network partitions lead to multiple active routers; these are mitigated through authentication mechanisms, such as MD5 in HSRP or simple text passwords in VRRP, which validate message authenticity and prevent unauthorized participation.[36] For IPv6 environments, FHRPs like HSRP for IPv6 and VRRPv3 integrate with Neighbor Discovery Protocol (NDP), where the active router sends unsolicited Neighbor Advertisements to update host neighbor caches during failover, replacing ARP equivalents. In Gateway Load Balancing Protocol (GLBP), failover maintains load distribution across multiple active virtual forwarders for continued traffic balancing.Comparisons and Applications
Feature and Performance Differences
The major first-hop redundancy protocols—Hot Standby Router Protocol (HSRP), Virtual Router Redundancy Protocol (VRRP), and Gateway Load Balancing Protocol (GLBP)—differ in their core features, with HSRP and GLBP being Cisco-proprietary while VRRP adheres to open standards defined in RFC 9568. HSRP operates on an active-standby model without native load balancing, using a single virtual MAC address per group and supporting multiple groups for manual load sharing across routers. VRRP employs a similar single-master election but allows only one active router per group, lacking built-in load distribution and focusing on transparent failover for IPv4 and IPv6 networks. In contrast, GLBP introduces load balancing through an Active Virtual Gateway (AVG) that assigns multiple Active Virtual Forwarders (AVFs), supporting up to four per group with weighted or round-robin algorithms to distribute traffic across gateways.[37][5][4] Performance metrics vary primarily in failover timing and scalability. HSRP and GLBP use default hello intervals of 3 seconds and hold times of 10 seconds, resulting in typical failover detection around 3-10 seconds, though subsecond convergence is achievable with Bidirectional Forwarding Detection (BFD) integration or timer tuning. VRRP defaults to a 1-second advertisement interval, enabling faster baseline failover of approximately 1-3 seconds, with Cisco implementations supporting millisecond timers for further optimization. Scalability limits include up to 4,000 HSRP groups per device in version 2, 255 VRRP groups per physical interface, and 1,024 GLBP groups per interface, all while maintaining low CPU overhead from periodic hello packets, typically under 1% utilization in standard deployments.[37][38][4] Security mechanisms emphasize authentication to prevent unauthorized participation. Both HSRP and VRRP support plain-text and MD5 authentication, with MD5 using key chains for enhanced protection against spoofing; however, VRRP version 3 per RFC 9568 (April 2024, obsoleting RFC 5798) removes all authentication support due to its limitations, relying instead on measures like a TTL/Hop Limit of 255 to mitigate remote injection attacks. GLBP offers similar MD5 and plain-text options but includes a strict mode that rejects hellos from non-GLBP devices, bolstering isolation in mixed environments. All protocols transmit via UDP multicast, with HSRP version 1 on 224.0.0.2, HSRP version 2 and GLBP on 224.0.0.102, and VRRP on 224.0.0.18.[37][5][4] Interoperability favors VRRP for multi-vendor setups due to its IETF standardization, enabling seamless integration across devices from different manufacturers. HSRP and GLBP, being Cisco-exclusive, limit deployment to Cisco ecosystems, potentially requiring protocol translation in heterogeneous networks.[5][37][4]| Attribute | HSRP | VRRP | GLBP |
|---|---|---|---|
| Standard | Cisco proprietary | IETF RFC 9568 (open standard) | Cisco proprietary |
| Load Balancing | No (manual via multiple groups) | No (single master) | Yes (weighted AVFs, round-robin, host-dependent) |
| Transport | UDP multicast (224.0.0.2 v1; 224.0.0.102 v2) | IP protocol 112 (multicast 224.0.0.18) | UDP multicast (224.0.0.102) |
| Default Failover | ~3-10 seconds (hello 3s, hold 10s) | ~1-3 seconds (advert 1s) | ~3-10 seconds (hello 3s, hold 10s) |
| Max Groups/Interface | Up to 4,000 per device (v2) | Up to 255 | Up to 1,024 |
| Authentication | Text, MD5 | None (authentication removed in RFC 9568) | Text, MD5 (strict mode for non-GLBP rejection) |
| Interoperability | Cisco-only | Multi-vendor | Cisco-only |