Jump to content


RIPv2


  • Please log in to reply
36 replies to this topic

#1 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 07 April 2009 - 06:13 AM

RIPv2

RIPv2 Summary
The characteristics of RIPv2 follow:

 Distance-vector protocol.

 Use UDP port 520.

 Makes Automatic summarization on network Class boundary.

 Classless protocol (supports VLSMs).

 Have the power to shut the Auto-summarization And make a configured manual Summarization.

 Supports VLSMs.

 Metric is router hop count.

 Maximum hop count is 15; infinite (unreachable) routes have a metric of 16.

 Periodic route updates sent every 30 seconds to multicast address 224.0.0.9.

 25 routes per RIP message (24 if authentication is used).

 Invalid route marked after 180 sec.

 Flush timer is 280 sec.

 Hold-down timer 180 sec.

 Supports authentication.

 Implements split horizon with poison reverse.

 Implements triggered updates.

 Subnet mask included in route entry.

 Administrative distance for RIPv2 is 120.

 Used in small, flat networks or at the edge of larger networks.


RIPv2 Convergence

The steps for RIPv2 convergence are as follows:

1- When the local router sees a connected route disappear, it sends a
Flash update and removes the route entry from its table. This is
Called a triggered update with poison reverse.

2- The receiving routers send a Flash update and put the affected route
in holddown.

3- The originating router queries its neighbor (DC) for alternative routes.
If the neighbor has an alternative route, it is sent; otherwise, the
Poisoned route is sent.

4- The originating router installs the best alternative route that it
hears because it has purged the original routes, and Send the new updates in
its periodic routing update timer (30 sec).

5- Routers that are in holddown ignore only the alternative route.

6- When the other routers emerge from holddown, they will accept
the alternative route.

7- Convergence takes the time for detection, plus holddown, plus the
number of routing updates (equal to the hop-count diameter of the network).
This could take a long time.



RIP v2 Configuration

config)# router rip
conf-router)# version 2
conf-router)# no auto-summary
conf-router)# network < classfull Net ID>

conf-router)#passive-interface <interface>

conf-router)#maximum –paths <N> (Up to 6 Links for load sharing on a R.B basis, Over equal metric paths default =is 4 paths)

conf-if)# no ip split-horizon (Disables it from a specific interface)
c
onf-router)#distance <N> (Define an administrative distance, default =120)

conf-router)# timers basic < update - invalid –holdown –flush – sleep > (Adjust timers in sec , Sleeptime is a timer used to specify a period, in milliseconds, to delay a regular routing update after receiving a triggered update.)

conf-router)# neighbor < ip add> (Because RIP is a broadcast protocol you may configure RIP to exchange routing information in a Point-Point non-broadcast network such as Frame Relay. In this type of network, RIP must be informed of neighbor RIP routers, Also When it is used in combination with the passive-interface cmd, routing information can be Only exchanged between the router and this nei Ex. Unix server on a LAN )

conf-router)#validate-update-source (This command is applicable only to RIP and IGRP. The software ensures that the source IP address of incoming routing updates is on the same IP network as one of the addresses defined for the receiving interface. Disabling split horizon on the incoming interface will also cause the system to perform this validation check. For unnumbered IP interfaces (interfaces configured as IP unnumbered), no checking is performed. )

conf-router)# no auto-summary
conf-if )# ip summary-address rip < Net ID + mask>


Show Commands

# sh ip route
# sh ip route rip
# sh ip route < Net ID >
# sh ip rip database
# sh ip protocols

#debug ip rip
#debug ip rip events
# no debug all


Hope this might be usefull

Edited by DarkFiber, 07 April 2009 - 06:15 AM.

  • 0

#2 jourjou

jourjou

    Newbie

  • Members
  • Pip
  • 16 posts
  • Gender:Male

Posted 07 April 2009 - 06:33 AM

thanks

it was a nice representation **
  • 1

#3 moekad

moekad

    Advanced Member

  • Members
  • PipPipPip
  • 270 posts

Posted 08 April 2009 - 07:06 AM

interesting
i just need about this:
you said in Router (config-router) # maximum-paths by default = 4 ( load balance if they got same AD and same metric), and you said beside up to 6 links ? i didn't got this point how ?
and Router(config-router)# validate-update-source you mean it make sure of the interface that recieve update can send update?
only i need to got answers of these 2 only
and thanks for the great summary and for your cooperation.
  • 0

#4 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 08 April 2009 - 08:03 PM

interesting
i just need about this:
you said in Router (config-router) # maximum-paths by default = 4 ( load balance if they got same AD and same metric), and you said beside up to 6 links ? i didn't got this point how ?
and Router(config-router)# validate-update-source you mean it make sure of the interface that recieve update can send update?
only i need to got answers of these 2 only
and thanks for the great summary and for your cooperation.


Dear Moekad,
The Max-paths cmd has a variable , you can config to have 4,5 or 6 equal metric paths , in New IOS you got the option to load balance over 16 equal metric paths...but you wont really see this in real world.

Validate update source cmd , it just confirm that your neighbor router which is directly connected within the same IP subnet.
  • 0

#5 YapChinHoong

YapChinHoong

    Advanced Member

  • Members
  • PipPipPip
  • 281 posts
  • Gender:Male

Posted 09 April 2009 - 12:06 AM

FYI, the maximum-paths support for RIP, OSPF, EIGRP, etc are 32 in the recent IOS releases.
  • 0

#6 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 09 April 2009 - 06:01 AM

FYI, the maximum-paths support for RIP, OSPF, EIGRP, etc are 32 in the recent IOS releases.


Really that's cool , last time i read was 16....I really wonder how even you will reach 5 paths : )
  • 0

#7 Shotokan

Shotokan

    Advanced Member

  • Members
  • PipPipPip
  • 199 posts

Posted 09 April 2009 - 07:02 AM

yeah... doesn't really see much of that in the real world now eh?
  • 0

#8 moekad

moekad

    Advanced Member

  • Members
  • PipPipPip
  • 270 posts

Posted 10 April 2009 - 01:40 AM

yeah... doesn't really see much of that in the real world now eh?


LOL 4 and it's enought to use ya your right :)
  • 0

#9 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 13 April 2009 - 08:29 PM

There was also a cmd that was deleted from the recent new IOS, which is traffic-share min , which allow you to have more than 1 path in the routing table but only 1 path that will be used for traffic,
so no load balance, you will just have for example 4 paths in the routing table , but your not making any load balance, This is only used to minimize the downtime delay, if this path goes down, you will be moved to the 2nd path with no delay as all paths are inserted in the routing table.....:)
  • 0

#10 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 30 April 2009 - 08:31 AM



Hello Guys , any questions regarding this Topic : )
  • 0

#11 testdude

testdude

    Advanced Member

  • Members
  • PipPipPip
  • 290 posts
  • Gender:Male
  • Location:Florida

Posted 30 April 2009 - 11:37 PM

QUOTE (DarkFiber @ Apr 29 2009, 06:31 PM) <{POST_SNAPBACK}>
Hello Guys , any questions regarding this Topic : )


Just one: do you have that in a Word doc? Like you did with the OSPF and EIGRP?

Thanks
  • 0

#12 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 01 May 2009 - 01:02 AM

QUOTE (testdude @ Apr 30 2009, 04:37 PM) <{POST_SNAPBACK}>
Just one: do you have that in a Word doc? Like you did with the OSPF and EIGRP?

Thanks


Here you go , enjoy my friend : )
Have Attached the Doc .

Tell me guys if you need any thing.

Attached Files


  • 0

#13 testdude

testdude

    Advanced Member

  • Members
  • PipPipPip
  • 290 posts
  • Gender:Male
  • Location:Florida

Posted 01 May 2009 - 03:44 AM

QUOTE (DarkFiber @ Apr 30 2009, 11:02 AM) <{POST_SNAPBACK}>
Here you go , enjoy my friend : )
Have Attached the Doc .

Tell me guys if you need any thing.


Very nice. Thank you!
  • 0

#14 dgharami

dgharami

    Newbie

  • Members
  • Pip
  • 22 posts
  • Gender:Male
  • Location:India

Posted 29 May 2009 - 07:53 PM

DarkFiber's RIPv2_CCNA in PDF Format

Attached File  RIPv2_CCNA.pdf   907.16KB   428 downloads

smile.gif
  • 0

#15 abhirushi

abhirushi

    Member

  • Members
  • PipPip
  • 60 posts
  • Gender:Male
  • Location:INDIA

Posted 11 June 2009 - 08:50 PM

QUOTE (DarkFiber @ Apr 30 2009, 04:01 AM) <{POST_SNAPBACK}>
Hello Guys , any questions regarding this Topic : )


Hi,
I have come across an issue while configuring RIP TIMERS (With timers-basic command) and also using default as well. I setup 3 routers (IOS 12.2 or above) and configured rip on them and when I shut 1 interface, the route for this network is removed from all routers’ routing table instantly where as according to rip timers it should use the process of invalid, hold down and flush timer. I connected all routers via Ethernet (Broadcast) media. I also checked rip database in which rip is sending triggered update by default. There is no option of disabling triggered updates on an interface (both ripv1 and rip v2). When I checked the documents for triggered update (Cisco sites), they are always using serial interfaces. So I don't know why in case of ethernet the timers are not coming into play and also what are the differences between rip functionality on ethernet /serial. Can you please look into this?

Thanking you in anticipation,

Regards
abhirushi



  • 0

#16 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 13 June 2009 - 04:20 AM

QUOTE (abhirushi @ Jun 11 2009, 01:50 PM) <{POST_SNAPBACK}>
Hi,
I have come across an issue while configuring RIP TIMERS (With timers-basic command) and also using default as well. I setup 3 routers (IOS 12.2 or above) and configured rip on them and when I shut 1 interface, the route for this network is removed from all routers’ routing table instantly where as according to rip timers it should use the process of invalid, hold down and flush timer. I connected all routers via Ethernet (Broadcast) media. I also checked rip database in which rip is sending triggered update by default. There is no option of disabling triggered updates on an interface (both ripv1 and rip v2). When I checked the documents for triggered update (Cisco sites), they are always using serial interfaces. So I don't know why in case of ethernet the timers are not coming into play and also what are the differences between rip functionality on ethernet /serial. Can you please look into this?

Thanking you in anticipation,

Regards
abhirushi


I really dont get your question, may you kindly say more details

Q 1
Q 2


Edited by DarkFiber, 13 June 2009 - 04:22 AM.

  • 0

#17 abhirushi

abhirushi

    Member

  • Members
  • PipPip
  • 60 posts
  • Gender:Male
  • Location:INDIA

Posted 13 June 2009 - 10:47 PM

QUOTE (DarkFiber @ Jun 12 2009, 11:50 PM) <{POST_SNAPBACK}>
I really dont get your question, may you kindly say more details

Q 1
Q 2


The topology is: R1 f0/0-------------------------f0/0 R2 f0/1-------------------------f0/1 R3
1)Two different subnets are configured between R1 R2 and R2 R3
2)Rip v2 is configured on all the routers. Able to ping from any to any.
3)Now I will shut the interface f0/0of R2, and immediately check the routing table on R3. The route 192.168.1.0/24 has disappeared instantly, instead of following the concept of timers in Rip (update, hold down, invalid, flush ).
When I went through rip database on the routers, triggered updates are enabled by default. As per cisco documentation, enabling/disabling triggered updates is possible only on serial interfaces.
So my questions are:
1) Why are the timers not coming into play on the ethernet interfaces (broadcast media)?
2) What are the differences between behavior of rip on ethernet and serial interfaces...?
Regards,
abhirushi
  • 0

#18 WassimZ

WassimZ

    Newbie

  • Members
  • Pip
  • 14 posts
  • Gender:Male
  • Location:Lebanon

Posted 03 July 2009 - 11:32 PM


Would someone be kind and explain to me why routers use load balancing?
And why you say that in real world we won't use more than 4 paths?

Thanks.
  • 0

#19 abhirushi

abhirushi

    Member

  • Members
  • PipPip
  • 60 posts
  • Gender:Male
  • Location:INDIA

Posted 03 July 2009 - 11:43 PM

QUOTE (WassimZ @ Jul 3 2009, 07:02 PM) <{POST_SNAPBACK}>
Would someone be kind and explain to me why routers use load balancing?
And why you say that in real world we won't use more than 4 paths?

Thanks.


A very detailed explanation on how& why load balancing works:
http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080094820.shtml
  • 0

#20 Asterix

Asterix

    Newbie

  • Members
  • Pip
  • 30 posts
  • Gender:Male

Posted 13 July 2009 - 03:47 AM

Thank you, great info!
  • 0

#21 ieee2

ieee2

    Newbie

  • Members
  • Pip
  • 8 posts

Posted 28 August 2009 - 05:36 PM

Thank you for the nice explanations.

I have a few fast question please. I did configure RIP V2 by using GNS3 as the following topology:
Posted Image

The thing that I could not understand is:
1) In the GNS manual, the changed the clock rate for every interface to 64000, why?

2) Is it a must to remove autosummary when configuring RIP V2 ?

3)If I am in the real world, How are the links is going to be established? I mean, the serial link between each router, if I have 3 branches for a company, and each branch is placed in a city, then how am I going to link the routers? please explain it as much as you can because I am not abl;e to iagine that.

Thank you.
  • 0

#22 TroyTempest

TroyTempest

    Member

  • Members
  • PipPip
  • 68 posts

Posted 29 August 2009 - 02:48 AM

Thank you for the nice explanations.

I have a few fast question please. I did configure RIP V2 by using GNS3 as the following topology:
Posted Image

The thing that I could not understand is:
1) In the GNS manual, the changed the clock rate for every interface to 64000, why?

2) Is it a must to remove autosummary when configuring RIP V2 ?

3)If I am in the real world, How are the links is going to be established? I mean, the serial link between each router, if I have 3 branches for a company, and each branch is placed in a city, then how am I going to link the routers? please explain it as much as you can because I am not abl;e to iagine that.

Thank you.



2) No, you don't have to remove autosummary with RIP v2.

Note that RIP Version 1 always uses automatic summarization. If you are using RIP Version 2, you can turn off automatic summarization by specifying no auto-summary. You would need to disable automatic summarization if you have to perform routing between disconnected subnets. When automatic summarization is off, subnets are advertised.

3) WAN links in the "real world" typically are provided by a Telco / Service Provider. They will physically install a copper or fibre connection from their network (usually from their closest exchange / PoP, depending on the technology) into your offices, and provide the DCE, to which you then connect your router serial interface. So for each office that you want to connect, the telco will need to bring in a cable and provide the interface for you to connect your router(s). They are responsible for transporting the data between your office routers over their back-haul network.

HTH
  • 0

#23 ieee2

ieee2

    Newbie

  • Members
  • Pip
  • 8 posts

Posted 29 August 2009 - 06:54 AM

2) No, you don't have to remove autosummary with RIP v2.

Note that RIP Version 1 always uses automatic summarization. If you are using RIP Version 2, you can turn off automatic summarization by specifying no auto-summary. You would need to disable automatic summarization if you have to perform routing between disconnected subnets. When automatic summarization is off, subnets are advertised.

3) WAN links in the "real world" typically are provided by a Telco / Service Provider. They will physically install a copper or fibre connection from their network (usually from their closest exchange / PoP, depending on the technology) into your offices, and provide the DCE, to which you then connect your router serial interface. So for each office that you want to connect, the telco will need to bring in a cable and provide the interface for you to connect your router(s). They are responsible for transporting the data between your office routers over their back-haul network.

HTH


Thank you.
2) Does this means, if I didn't use autosummary in RIP v2, then it will act like RIP v1 (in the point where it doesn't sdvertise subnets?)

I am still wondering, why do I have to change the clock rate to 64000 for each serial interface when configuring RIP v2. (This is what I found in GNS3 manual, and I don't know why they did that, should all serial interfaces have the same clock rate??).

Thank you.
  • 0

#24 TroyTempest

TroyTempest

    Member

  • Members
  • PipPip
  • 68 posts

Posted 29 August 2009 - 08:13 PM

Thank you.
2) Does this means, if I didn't use autosummary in RIP v2, then it will act like RIP v1 (in the point where it doesn't sdvertise subnets?)

I am still wondering, why do I have to change the clock rate to 64000 for each serial interface when configuring RIP v2. (This is what I found in GNS3 manual, and I don't know why they did that, should all serial interfaces have the same clock rate??).

Thank you.


2) Yes.

You only need to set the clock rate on a DCE serial interface - but if all your serial interfaces are DCE then you would need to set the clock rate on all of them. The Clock Rate is set in bps and should be set to the actual link speed - this is normally done by the link provider, btw. I don't know GNS3, so perhaps there is something in there that only supports 64K!

RIP doesn't actually know or care about the actual link speed, as it doesn't use it in determining routes (unlike OSPF and EIGRP), so I'm surprised that there's a connection between RIP and bandwidth in the GNS3 manual.

TT
  • 0

#25 VENERZKY

VENERZKY

    Advanced Member

  • Members
  • PipPipPip
  • 151 posts
  • Gender:Male
  • Location:Makati City, Philippines

Posted 08 September 2009 - 05:11 AM

GO!!!!!!!!

Edited by VENERZKY, 08 September 2009 - 05:20 AM.

  • 0

#26 alex1966

alex1966

    Newbie

  • Members
  • Pip
  • 1 posts

Posted 08 October 2009 - 02:26 AM

Thanks, great wrap up. One question though: how did you get that flash time is 280 sec.? Always thought it was 240 sec.
Thanks. :unsure:
  • 0

#27 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 09 October 2009 - 12:43 AM

Thanks, great wrap up. One question though: how did you get that flash time is 280 sec.? Always thought it was 240 sec.
Thanks. :unsure:



OH Nice catch buddy, will make an edit on that :)
  • 0

#28 Mr.Mandeep

Mr.Mandeep

    Newbie

  • Members
  • Pip
  • 14 posts

Posted 20 December 2009 - 12:32 AM

RIPv2



RIPv2 Convergence

The steps for RIPv2 convergence are as follows:

1- When the local router sees a connected route disappear, it sends a
Flash update and removes the route entry from its table. This is
Called a triggered update with poison reverse.

2- The receiving routers send a Flash update and put the affected route
in holddown.




Okay. Let's start with your step 1.


A router with a network down will send a triggered update with poisoning the route, thereby flushing the route completely from routing table. Fair enough

In your step second, You mentioned " Receiving routers send a flash update and put the affected route in holdown".

Isn't it A route which is flushed from routing table will never be put in holdown?

There is no holdown interval for a route that has been flushed.
  • 0

#29 DarkFiber

DarkFiber

    Cisco Expert

  • Members
  • PipPipPipPip
  • 927 posts
  • Gender:Male
  • Location:Cairo-Egypt
  • Interests:Networks

Posted 20 December 2009 - 09:53 PM

Okay. Let's start with your step 1.


A router with a network down will send a triggered update with poisoning the route, thereby flushing the route completely from routing table. Fair enough

In your step second, You mentioned " Receiving routers send a flash update and put the affected route in holdown".

Isn't it A route which is flushed from routing table will never be put in holdown?

There is no holdown interval for a route that has been flushed.


The 1st router will Flush the route & send a FLASH update
other routers will receive the FLASH update & put the route into the (holdown, invalid) timer , and then FWD the FLASH update

Hope this made it more clear
  • 0

#30 Mr.Mandeep

Mr.Mandeep

    Newbie

  • Members
  • Pip
  • 14 posts

Posted 21 December 2009 - 12:16 AM

The 1st router will Flush the route & send a FLASH update
other routers will receive the FLASH update & put the route into the (holdown, invalid) timer , and then FWD the FLASH update

Hope this made it more clear



This is incorrect. RIP timers are poorly explained in documentations.



As I said before based on experiments,

The route will go into holdown for a route that is missing, means it was not confirmed. During holdown period, any updates about that route with better or worse metric are ignored.


Whenever an update confirming a network is received, the invalid-after timer is reset to 0. If it gets up to 180 seconds by default, the route goes into holdown.


For a route that is flushed, means, it wil be removed from the routing table. No holddown for that route.


I hope, this helps.
  • 0





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users