RIP is the oldest routing protocol most candidates meet, and it is the cleanest way to learn how distance-vector routing behaves: routers know only what their neighbors tell them, and the problems that follow (slow convergence, loops, count to infinity) are the reasons newer protocols like OSPF and EIGRP exist. This guide covers how RIP works, the three versions, the timers, loop prevention, a working Cisco IOS configuration and where it fits in the CCNA picture.
How does the RIP routing protocol work?
Each RIP router starts knowing only its directly connected networks. Every 30 seconds it sends its routing table to its neighbors in a RIP response message over UDP port 520. A neighbor that receives a route adds 1 to the metric (the cost of crossing the link), compares it with what it already has, and keeps the lower hop count. After a few rounds of updates every router knows a path to every network, and each routing table entry records the next hop and the number of routers to cross.
A RIP router also sends a request at startup to ask neighbors for their full tables immediately, and sends triggered updates as soon as a metric changes rather than waiting for the next 30-second tick. A single RIP message carries up to 25 routes, so large tables span many messages. RIP is called “routing by rumor” because a router trusts its neighbor's word about the rest of the network and never sees the full topology, which is the defining difference from the link-state design used by OSPF.
Equal-cost paths are load-balanced. On Cisco IOS RIP installs up to four equal-hop paths by default (adjustable with maximum-paths). Because the metric ignores bandwidth, a path across one slow serial link and a path across one gigabit link look identical.
RIPv1 vs RIPv2 vs RIPng: what is the difference?
RIPv1 is classful and broadcasts updates without subnet masks. RIPv2 is classless, sends the mask and next hop, and uses multicast 224.0.0.9. RIPng is the separate IPv6 version, using FF02::9 and UDP port 521. All three use hop count and a 15-hop limit.
| Feature | RIPv1 | RIPv2 | RIPng |
|---|---|---|---|
| Defined in | RFC 1058 | RFC 2453 | RFC 2080 |
| Address family | IPv4 | IPv4 | IPv6 |
| Classful or classless | Classful | Classless | Classless (prefix length in every route) |
| Subnet mask in updates | No | Yes | Yes (prefix length) |
| VLSM and discontiguous networks | Not supported | Supported (disable auto-summary) | Supported |
| Update destination | Broadcast 255.255.255.255 | Multicast 224.0.0.9 | Multicast FF02::9 |
| Authentication | None | Plain text or MD5 (see RFC 2453, RFC 4822) | None built in; relies on IPsec |
| Transport | UDP 520 | UDP 520 | UDP 521 |
| Metric and limit | Hop count, 15 max | Hop count, 15 max | Hop count, 15 max |
| Administrative distance (Cisco) | 120 | 120 | 120 |
Classful versus classless is the distinction that matters most in the exam. A classful protocol assumes the mask from the address class (Class A /8, B /16, C /24), so it can only handle one mask per network and cannot do VLSM. A classless protocol sends the mask with every route. RIPv2 added the mask, a next-hop field, route tags and authentication while keeping the rest of RIPv1's behavior, which is why it is backward compatible in message format. The IPv6 version, RIPng, is a separate protocol with a new message format and the same distance-vector logic.
What are the RIP timers?
RIP uses four timers. On Cisco IOS the defaults are update 30 seconds, invalid 180 seconds, holddown 180 seconds and flush 240 seconds. These are the values Cisco exams and show commands reflect:
| Timer | Default | What it does |
|---|---|---|
| Update | 30 seconds | How often the router sends its routing table to neighbors. |
| Invalid (timeout) | 180 seconds | If no update for a route arrives in this time, the route is marked unreachable (metric 16) but still advertised so neighbors learn it is down. |
| Holddown | 180 seconds | After a route is declared unreachable, the router ignores updates that offer a worse or equal path for that network until the timer ends, to avoid being fooled by stale information. |
| Flush | 240 seconds | Measured from the last valid update. When it expires the route is removed from the routing table (60 seconds after it went invalid). |
Two caveats worth knowing. First, the numbers above describe Cisco IOS. RFC 2453 specifies a 180-second timeout and then a separate 120-second garbage-collection period, and it does not define a holddown timer; holddown is a Cisco addition, so other vendors differ. Second, these can be changed with timers basic update invalid holddown flush, but timers must match on every router in the domain or convergence becomes unpredictable. With defaults, a dead route can take several minutes to disappear from a distant router's table, which is the slow convergence RIP is known for.
How does RIP prevent routing loops?
Distance-vector protocols can create loops: A believes it can reach a network via B, B believes it via A, and a packet bounces between them. After a failure, the metrics increase on each update round, which is called count to infinity. RIP combines several defenses, and exam questions often ask you to name the right one:
| Mechanism | What it does |
|---|---|
| Maximum hop count (16 = infinity) | Caps count to infinity: after at most 16 increments the route is dead. Also limits network diameter to 15 hops. |
| Split horizon | Never advertise a route out of the interface it was learned on. Stops the simplest two-router loop. |
| Route poisoning | When a route fails, advertise it with metric 16 so neighbors remove it immediately instead of waiting for timers. |
| Poison reverse (split horizon with poisoned reverse) | Instead of staying silent about a route on the interface it came from, advertise it there with metric 16 so the neighbor knows for certain not to use you as a path. |
| Holddown timer | Ignores worse information about a failed route for 180 seconds while the network settles. |
| Triggered updates | Send a change immediately instead of waiting for the next 30-second update. |
Keep the terminology straight: split horizon is silence, poison reverse is an explicit “unreachable” message on that same interface, and route poisoning is advertising a failed route as unreachable. Cisco IOS enables split horizon by default on most interfaces, and uses poisoning when routes fail. On a multipoint interface such as hub-and-spoke WAN links, split horizon can stop spokes learning each other's routes, a classic design trap.
How do you configure RIP on Cisco IOS?
Enter router rip, set version 2, add no auto-summary, and list a classful network statement for each directly connected network. Consider three routers in a line: R1 (LAN 10.1.1.0/24) connects over 192.168.12.0/24 to R2, which connects over 192.168.23.0/24 to R3 (LAN 10.3.3.0/24). This is the configuration for R1. R2 and R3 are the same pattern with their own networks.
R1(config)# router rip
R1(config-router)# version 2
R1(config-router)# no auto-summary
R1(config-router)# network 10.0.0.0
R1(config-router)# network 192.168.12.0
R1(config-router)# passive-interface GigabitEthernet0/0Points to notice. network statements are classful: you enter the classful network (10.0.0.0, 192.168.12.0), and RIP enables itself on every interface in that network, even though RIPv2 is classless in what it sends. version 2 forces RIPv2 for sending and receiving; without it a Cisco router sends RIPv1 updates but accepts both. no auto-summary stops RIP summarizing at classful boundaries, which is needed for discontiguous subnets and is a standard exam line. passive-interface stops updates on the LAN interface (no routers there to hear them) while still advertising its network. Add default-information originate on an edge router to advertise a default route, and MD5 authentication uses key chains with ip rip authentication mode md5.
For RIPng on IPv6, you enable it per interface rather than with network statements:
R1(config)# ipv6 unicast-routing
R1(config)# ipv6 router rip LAB
R1(config)# interface GigabitEthernet0/1
R1(config-if)# ipv6 address 2001:db8:12::1/64
R1(config-if)# ipv6 rip LAB enablePractice routing questions free
Try 30 CCNA questions per session in free practice mode, no card needed. The full 500-question bank and exam mode unlock with a paid plan.
Check yourself
Check 1. What metric does RIP use, and what value means unreachable?
Hop count. A metric of 16 means unreachable; the longest usable path is 15 hops.
Check 2. Which RIP version sends updates to 224.0.0.9 and includes the subnet mask?
RIPv2. RIPv1 broadcasts to 255.255.255.255 and carries no mask.
Check 3. What is the default administrative distance of RIP on Cisco routers?
120.
How do you verify and troubleshoot RIP?
Three commands cover almost everything. On R1 in the topology above, the output would look like this (illustrative):
R1# show ip route rip
R 192.168.23.0/24 [120/1] via 192.168.12.2, 00:00:14, GigabitEthernet0/1
R 10.3.3.0/24 [120/2] via 192.168.12.2, 00:00:14, GigabitEthernet0/1
R1# show ip rip database
10.1.1.0/24 directly connected, GigabitEthernet0/0
10.3.3.0/24
[2] via 192.168.12.2, 00:00:14, GigabitEthernet0/1
192.168.12.0/24 directly connected, GigabitEthernet0/1
192.168.23.0/24
[1] via 192.168.12.2, 00:00:14, GigabitEthernet0/1Read the routing table entry [120/2] as administrative distance 120 and metric 2 hops. The 00:00:14 is the time since the last update, which should reset to zero every 30 seconds or so; if it keeps climbing past 30, updates are not arriving. show ip protocols prints the version, the timers, the networks being advertised and the neighbors (“Routing Information Sources”). debug ip rip shows updates as they are sent and received, but use it briefly on a production device.
The usual faults, in the order to check them:
- Missing or wrong
networkstatement. RIP will not run on an interface whose network is not covered. - Passive interface on a link that needs neighbors. Updates are suppressed, so no adjacency-like exchange happens.
- Version mismatch. A router sending only RIPv1 to a router configured for RIPv2 only will be ignored. Set the version explicitly on every router.
- Auto-summary with discontiguous networks. Two sites advertise the same classful summary, so one side loses traffic. Use
no auto-summary. - Hop count over 15. Remote networks silently never appear.
- ACL or layer 2 problem. RIPv2 needs UDP 520 and the multicast 224.0.0.9 to pass, and the interface must be up/up.
Why is RIP rarely used today?
RIP is rarely chosen for new networks because its hop-count metric ignores link speed, it cannot span more than 15 routers, and it converges slowly. The main limitations are:
- Hop count ignores bandwidth. A two-hop gigabit path loses to a one-hop 56 kbps link.
- 15-hop limit. Large or deep networks cannot use it.
- Slow convergence. Timers are measured in minutes, versus seconds for OSPF or EIGRP.
- Periodic full-table updates. Every router resends everything every 30 seconds, wasting bandwidth on large tables.
- Weak security. RIPv1 has no authentication, and RIPv2 authentication is optional and often skipped.
- Limited tuning. No areas, no real route tagging policy compared with modern protocols.
It remains reasonable for tiny, stable networks, labs, and some embedded or legacy devices where simplicity matters more than speed. For a branch that needs fast failover, or a campus above a handful of routers, OSPF or EIGRP is the answer.
RIP vs OSPF vs EIGRP: how do they compare?
RIP is a simple distance-vector protocol with a 15-hop limit and AD 120. OSPF is link-state with AD 110 and scales with areas. EIGRP is an advanced distance-vector protocol with AD 90 for internal routes and the fastest convergence of the three.
| Feature | RIP | OSPF | EIGRP |
|---|---|---|---|
| Algorithm / type | Distance vector (Bellman-Ford) | Link state (Dijkstra SPF) | Advanced distance vector (DUAL) |
| Metric | Hop count | Cost (based on bandwidth) | Composite: bandwidth and delay by default |
| Administrative distance (Cisco) | 120 | 110 | 90 internal / 170 external |
| Max network size | 15 hops | Very large, with areas | Large (default maximum hop count of 100 is configurable) |
| Convergence | Slow (timer-driven) | Fast | Very fast |
| Updates | Full table every 30 s | Triggered LSAs, periodic refresh every 30 minutes | Partial, triggered updates after neighbor adjacency |
| Transport | UDP 520 | IP protocol 89 | IP protocol 88 |
| Multicast address (IPv4) | 224.0.0.9 (v2) | 224.0.0.5 / 224.0.0.6 | 224.0.0.10 |
| Standard | Open (RFC 1058, 2453) | Open (RFC 2328) | Originally Cisco; basic functionality published as RFC 7868 |
| Typical use | Tiny networks, labs, legacy | Enterprise and service provider | Cisco-centric enterprise |
For the full comparison of the two protocols that replaced RIP in most designs, read our OSPF and EIGRP guides. Remember that administrative distance chooses between protocols and metric chooses within one protocol.
What is the exam angle?
RIP is background knowledge rather than a configuration topic: expect it as a distance-vector example, an administrative distance of 120, or a distractor next to OSPF. The details below should be checked against the current objectives.
Cisco CCNA 200-301 centres its dynamic routing objectives on OSPFv2 (single-area configuration), along with static routing, the routing table, administrative distance, metric and longest prefix match. RIP is not a configuration objective; expect it as a distractor or as background in questions that compare distance-vector and link-state protocols, ask for the administrative distance, or use RIP output to test routing table reading. CompTIA Network+ N10-009 focuses on the routing concepts and protocols (OSPF, EIGRP and BGP), so RIP knowledge is still useful for understanding distance-vector and for older study material. Objectives change between versions, so confirm against the official CCNA exam topics and the CompTIA Network+ page.
High-value facts to memorize:
- Hop count, maximum 15, 16 = unreachable. Administrative distance 120.
- RIPv1 classful broadcast; RIPv2 classless multicast 224.0.0.9; RIPng IPv6 FF02::9 on UDP 521.
- Update 30, invalid 180, holddown 180, flush 240 (Cisco defaults).
- Split horizon, route poisoning, poison reverse, holddown and triggered updates are the loop-prevention set.
no auto-summaryandversion 2in nearly every RIPv2 config.
Related Layer 2 and Layer 3 loop-prevention topics sit nearby in the blueprint: Spanning Tree Protocol stops loops in switched networks, and HSRP gives hosts a redundant default gateway. For a full topic map, see the CCNA cheat sheet.
RIP practice questions with answers
Answer first, then open each one. For more exam-style questions use the CCNA practice exam or read our CCNA practice questions.
1. A network uses 10.1.0.0/24 and 10.2.0.0/16 subnets with different masks and runs RIPv1. What goes wrong?
RIPv1 is classful and sends no mask, so it cannot carry variable-length subnet masks. The routers will misinterpret or ignore the routes. Switch to RIPv2 (or OSPF or EIGRP), which carry the mask in every update.
2. A route is learned with metric 15. A neighbor advertises it, and the receiving router adds one hop. What happens?
The metric becomes 16, which RIP treats as infinity, so the route is considered unreachable and is not installed. This is the 15-hop limit in action.
3. Which feature stops a router advertising a route back out of the interface it was learned on?
Split horizon. It prevents the simplest two-router routing loops. Route poisoning marks a failed route as metric 16, and poison reverse (split horizon with poisoned reverse) advertises a learned route back with metric 16 instead of omitting it. Holddown and triggered updates are separate mechanisms.
4. A router has not heard an update for a RIP route in 185 seconds with default timers. What state is the route in?
Invalid (also described as possibly down). The invalid timer is 180 seconds, so after it expires the route is marked with metric 16 and enters holddown; it is removed when the flush timer reaches 240 seconds from the last update.
5. A router learns 172.16.5.0/24 from RIP and from OSPF. Which is installed?
The OSPF route, because OSPF has an administrative distance of 110 and RIP has 120. Administrative distance is compared first, between protocols; metric is only compared within a single protocol.
6. Two routers run RIPv2 with auto-summary on. They are separated by another network and each has subnets of 172.16.0.0/16. Why does connectivity break?
Each router summarizes its subnets to the classful boundary 172.16.0.0/16 when advertising across a different classful network, so both advertise the same summary and the routers cannot tell which side holds which subnet. This is the discontiguous network problem. The fix is no auto-summary.
7. Which command shows the routes RIP has learned, and which shows the RIP database?
show ip route rip lists RIP routes in the routing table (marked R). show ip rip database lists everything RIP knows, including summary and directly connected entries. show ip protocols shows the timers, version and networks.
8. Which statement describes RIPng?
It is the IPv6 version of RIP (RFC 2080), using UDP port 521 and the multicast address FF02::9. RIPng is a different protocol from RIPv2 rather than an extension of it, and it keeps the 15-hop limit.
Frequently asked questions
What is the RIP routing protocol?
RIP (Routing Information Protocol) is a distance-vector interior routing protocol that uses hop count as its metric. Each router periodically shares its routing table with its neighbors, and a route is chosen by the lowest number of routers to cross. RIP allows a maximum of 15 hops; a metric of 16 means the destination is unreachable. RIPv1 is defined in RFC 1058 and RIPv2 in RFC 2453.
What is the difference between RIPv1 and RIPv2?
RIPv1 is classful: it sends no subnet mask, broadcasts updates to 255.255.255.255 and has no authentication, so it cannot support VLSM or discontiguous networks. RIPv2 is classless: it includes the subnet mask and next hop in each route, sends updates to the multicast address 224.0.0.9, and supports authentication. Both use UDP port 520, a hop-count metric and an administrative distance of 120.
What is RIPng?
RIPng (RIP next generation) is the IPv6 version of RIP, defined in RFC 2080. It carries IPv6 prefixes and prefix lengths, sends updates to the multicast group FF02::9 and uses UDP port 521. It keeps the same 15-hop limit and the same distance-vector behavior as RIPv2.
What are the RIP timers?
On Cisco IOS the defaults are: update 30 seconds, invalid 180 seconds, holddown 180 seconds and flush 240 seconds. The update timer controls how often the full table is sent; a route not refreshed within the invalid timer is marked unreachable (metric 16); holddown stops the router accepting worse information about it; and the flush timer, measured from the last update, removes the route. RFC 2453 itself defines a 180-second timeout and a 120-second garbage-collection timer, and does not define holddown.
Why is RIP limited to 15 hops?
RIP treats a metric of 16 as infinity, which gives the protocol a way to stop count-to-infinity loops: if a loop forms, the metric climbs by one on each update until it reaches 16 and the route is declared unreachable. The trade-off is that no destination can be more than 15 routers away, which makes RIP unsuitable for large networks.
What is the administrative distance of RIP?
On Cisco routers RIP has an administrative distance of 120. This is higher (less trusted) than EIGRP internal routes (90) and OSPF (110), and lower than external EIGRP (170) and iBGP (200). When two protocols offer a route to the same prefix, the lower administrative distance wins.
Is RIP still used, and is it on the CCNA or Network+ exam?
RIP is rarely chosen for new designs but still appears in small or legacy networks, labs and some embedded devices. In recent exam versions it is mostly a concept to recognize: distance-vector behavior, hop count, administrative distance 120. Configuration of OSPFv2 is the named dynamic-routing objective in Cisco CCNA 200-301. Always check the current official objectives for the exact wording.
How we wrote and checked this
CertSharp is a practice-exam provider, so we have a commercial interest in exam preparation. This guide was built from RFC 1058, RFC 2453 and RFC 2080 and Cisco documentation. The configuration examples follow Cisco IOS syntax from vendor documentation, and the show output is illustrative rather than captured from production equipment. Timers, addresses and protocol numbers were cross-checked against the RFCs. Last reviewed on October 18, 2026. Confirm exam objective wording on the Cisco and CompTIA pages.
Sources and further reading
- RFC 1058: Routing Information Protocol
- RFC 2453: RIP Version 2
- RFC 2080: RIPng for IPv6
- RFC 4822: RIPv2 Cryptographic Authentication
- RFC 2328: OSPF Version 2
- RFC 7868: Cisco's Enhanced Interior Gateway Routing Protocol (EIGRP)
- Cisco Learning Network: CCNA 200-301 exam topics
- CompTIA Network+ certification page
Practice CCNA routing questions
30 free questions per session in practice mode, no signup and no credit card. The full 500-question bank with exam mode is $11.99 lifetime per certification, or $9.99/month Pro unlocks every CertSharp certification.