Virtual router redudancy protocol questions

Answered Question
Jan 22nd, 2009

HI every body!

I have few questions:

1)In HSRP, one router is chosen as active while the other is chosen as standby, all the rest would be in listening state. What about VRRP?

In vrrp, one router is chosen as master,are all the rest chosen as back up routers? or only one router among them is chosen as backup router?

2)In HSRP, active router overwrite the timers configured on other routers in the group. How about in VRRP? does master in vrrp overwrite the timers on other routers in the group?

3) which routers can receive and send hello in VRRP and which ones can only listen ?

4) Do routers use hold time in VRRP?

thanks a lot!

I have this problem too.
0 votes
Correct Answer by lejoe.thomas about 7 years 10 months ago

Hi Sarah

1) In VRRP, one router becomes the Master (Virtual Router Master) and others act as backup (Virtual Router Backup) like HSRP. Eg: Lets assume three routers R1,R2,R3. R1 is currently the master. If R1 goes down, then R2 and R3 depending on their priority or if the priorities are same, depending on the higher IP address become the Master.

2) Well there are two ways in HSRP,

vrrp group timers advertise [msec] interval

Configures the interval at which Master sends advertisement. If you only use the above command, then you need to set timers to be consistent on all other routers in the group. Default interval is 1 sec

vrrp group timers learn

If this command is configured, then it informs all backup routers to learn the advertisement interval from the Master.

For consistency sake, you configure the above two commands to be same on all Routers in the VRRP group.

3) Virtual Router Master sends out (hellos) multicast advertisement, every 1 sec by default and backups listen.

4) Well I'd assume it's roughly 3 hellos. The example in the paper below shows its to be roughly 3.6 or 3.5 secs if the advertisement interval 1 sec.

http://www.cisco.com/en/US/docs/ios/ipapp/configuration/guide/ipapp_vrrp_ps6350_TSD_Products_Configuration_Guide_Chapter.html#wp1054492

HTH

Lejoe

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 4 (4 ratings)
Loading.
Correct Answer
lejoe.thomas Thu, 01/22/2009 - 21:04

Hi Sarah

1) In VRRP, one router becomes the Master (Virtual Router Master) and others act as backup (Virtual Router Backup) like HSRP. Eg: Lets assume three routers R1,R2,R3. R1 is currently the master. If R1 goes down, then R2 and R3 depending on their priority or if the priorities are same, depending on the higher IP address become the Master.

2) Well there are two ways in HSRP,

vrrp group timers advertise [msec] interval

Configures the interval at which Master sends advertisement. If you only use the above command, then you need to set timers to be consistent on all other routers in the group. Default interval is 1 sec

vrrp group timers learn

If this command is configured, then it informs all backup routers to learn the advertisement interval from the Master.

For consistency sake, you configure the above two commands to be same on all Routers in the VRRP group.

3) Virtual Router Master sends out (hellos) multicast advertisement, every 1 sec by default and backups listen.

4) Well I'd assume it's roughly 3 hellos. The example in the paper below shows its to be roughly 3.6 or 3.5 secs if the advertisement interval 1 sec.

http://www.cisco.com/en/US/docs/ios/ipapp/configuration/guide/ipapp_vrrp_ps6350_TSD_Products_Configuration_Guide_Chapter.html#wp1054492

HTH

Lejoe

sarahr202 Fri, 01/23/2009 - 08:22

Hi Lejoe!

"vrrp group timers advertise [msec] interval "

Since only master send hellos, the above command is only relevant on master. Is it correct ?

"vrrp group timers learn "

The above command causes back up routers to learn timers value from master.As you mentioned ,the back up routers only listen advertisements from master, the only reason i could think of, for backup routers to learn timers values from master, is to use these timers, should one of them become a master.

Is it correct or wrong?

thanks a lot Lejoe and have a nice day!

lejoe.thomas Fri, 01/23/2009 - 14:36

Hi Sarah,

vrrp group timers advertise [msec] interval

The above must be configured on all routers in the VRRP group (including master and backup).

Let me demonstrate this with a simple example

3640a is Virtual Master for VRRP group 1, timers have been changed from default to 2 sec. Hence the holdtime will roughly be 6secs. Priority is 101.

3640a#sh run int e0/1

interface Ethernet0/1

ip address 192.168.0.1 255.255.255.0

vrrp 1 ip 192.168.0.254

vrrp 1 timers advertise 2

vrrp 1 priority 101

2610xm is Virtual Router Backup for VRRP group 1. Timers havent been changed and it isnt configured to learn timers from Master. It assumes advertisement interval is 1 sec and holdtime is roughly 3 secs. Default priority is 100.

2610XM#sh run int f0/0

interface FastEthernet0/0

ip address 192.168.0.2 255.255.255.0

duplex auto

speed auto

vrrp 1 ip 192.168.0.254

end

Because we havent configured the timers to be consistent on both routers (master and backup), Both routers going to assume master status.

Supposed to be master

3640a#sh vrrp all

Ethernet0/1 - Group 1

State is Master

Virtual IP address is 192.168.0.254

Virtual MAC address is 0000.5e00.0101

Advertisement interval is 2.000 sec

Preemption enabled

Priority is 101

Master Router is 192.168.0.1 (local), priority is 101

Master Advertisement interval is 2.000 sec

Master Down interval is 6.605 sec

Even the virtual backup thinks it's the Master because of timer mismatch.

supposed to be backup, but Master

2610XM#sh vrrp all

FastEthernet0/0 - Group 1

State is Master

Virtual IP address is 192.168.0.254

Virtual MAC address is 0000.5e00.0101

Advertisement interval is 1.000 sec

Preemption enabled

Priority is 100

Master Router is 192.168.0.2 (local), priority is 100

Master Advertisement interval is 1.000 sec

Master Down interval is 3.609 sec

To resolve this, we need to configure

vrrp 1 timers advertise 2 or vrrp 1 timers learn on the virtual backup router as well (2610XM)

I'll demonstrate by adding vrrp 1 timers advertise 2 on the backup (2610xm), so timers are consistent with the master.

2610XM(config)#int f0/0

2610XM(config-if)#vrrp 1 timers advertise 2

2610XM#sh

*Mar 1 00:29:53.528: %VRRP-6-STATECHANGE: Fa0/0 Grp 1 state Master -> Backup

2610XM#sh vrrp all

FastEthernet0/0 - Group 1

State is Backup

Virtual IP address is 192.168.0.254

Virtual MAC address is 0000.5e00.0101

Advertisement interval is 2.000 sec

Preemption enabled

Priority is 100

Master Router is 192.168.0.1, priority is 101

Master Advertisement interval is 2.000 sec

Master Down interval is 6.609 sec (expires in 5.455 sec)

As you can see, 2610xm has finally accepted 3640a as the Master and itself as the backup.

vrrp group timers learn

It too should be configured on all Routers. This is because if Master Goes down and pre-emption is disabled and it comes back online. We would it want it to learn the timers from the backup routers promoted to Master status.

In summary, use both commands (timers learn and advertise) and configure it on all routers in the VRRP group.

HTH

Lejoe

franktuck Fri, 01/23/2009 - 10:09

Harold,

Thanks for the info. In the current configuration I'm doing rip, but I'm not sure if hsrp on the Fe interface will affect rip.

Can you comment on this?

sarahr202 Fri, 01/23/2009 - 12:00

Hi Frank!

first of all the subnet on interface (Fe) must be advertised by " network " command. For that, you must configure the network 199.199.199.0(say you have 199.199.199.0 subnet on Fe) under rip. we need to do that so other routers participating in rip will know how to reach 199.199.199.0 subnet.

Configuring hsrp has no effect on rip protocol.

Please feel free if you need any more help( you must see me the way I bug these nice people here on net pro until I find my answer).

franktuck Fri, 01/23/2009 - 14:42

Hi Sarah,

Thanks.

In my HSRP configuration I have two 2811, each with a serial interface and fastethernet interface. The HSRP will be configured on the fastethernet.

Q: I believe both routers will receive the same data on serial side, however will the standby router pass on its data to the fastethernet

sarahr202 Fri, 01/23/2009 - 18:53

Hi Frank!

I need some more detail( excuse my low IQ :-)

" I believe both routers will receive the same data on serial side"

what do you mean by data here?

There are two kinds of traffic among others here which are important in our scenario:

1) routing updates

2) user traffic

If you mean rip updates by"data" since you are running rip on your routers then consider the following.

let say we have two 2811 routers, R1 and R2.

Assume Fe on both routers are on 199.199.199.0

Let assume you are advertising this network in rip i.e network 199.199.199.0 on both routers.

R1 will advertise network 199.199.199.0 via rip updates on its serial interface connected towards R2 . R2 receives the updates but ignores it. Why? because the Ad for rip is 120 while Ad for connected network which is 199.199.199.0 is zero. Similarly R2 advertise the 199.199.199.0 in rip updates on its serial interface towards R1. R1 will ignore the update because the AD for rip 120 while AD for connected network is 0.

If you mean user-Ip -packet on the network, then consider the following.

R1 is active while R2 is standby.

R2 receives an ip packet destined to 199.199.199.4( say some host connected to same network i.e 199.199.199.0).

R2 checks its routing table and find entry which would be" 199.199.199.0 directly connected fe"

R2 simply forwards the packet out of its FE.

If I did not explain it good, come back, i will try harder next time :-)

franktuck Fri, 01/23/2009 - 19:09

Hi Sarah,

I think I'm good with the rip explanation. However, will the user traffic sent by the standby router R2 cause any problems?

sarahr202 Fri, 01/23/2009 - 20:11

No though it will result in asymmetric routing.

Because R2 forwards the packet destined to 199.199.199.4 out of FE( let say Fe is on 199.199.199.0 ). If host (199.199.199.4) has to send the packet, it will be forwarded to R1(active) .

The end result is it will cause asymmetric routing which is when packet from the same source to destination and vice versa does follow the same path.

It has no bearing on your serial interface.

Come back if you need more help.

franktuck Fri, 01/23/2009 - 19:39

Hi Sarah,

Let me provide you with some more detail.

In the current configuration R1's serial interface is a VWIC-2MFT-T1 card, and the connection cabel between the router and the microwave/WAN is RJ45 on one end and wire wrapped at the other end.

I'm proposing that the R2 router would have the same connection cable and wire wrapped to the same pins.

Q: Will this affect the rip updates as you have described?

I hope this has shead some more light on this subject.

Thanks.

sarahr202 Fri, 01/23/2009 - 20:02

Sorry Frank English is my secondary language :-)

I still don't get your topology.

Could you rephrase your topology please?

Let me try what i understand

r1s0-----------rj45(microwave wan)rj45---s0r2

Did i understand you correct?

If i did, then i could simply redraw your topology as

r1s0----------------------s0r2

franktuck Sat, 01/24/2009 - 17:07

Hi Sarah,

I have attached a drawing of the network.

The router (red) and the one below will be connected to a patch panel. I'm not sure if that connection will be on the same pins.

If they are connected to the same pins will this affect the rip updates between the R1 & R2? (R2 is standby)

sarahr202 Sun, 01/25/2009 - 15:43

Hi Frank!

you said router 2 is standby which it can not be.

The router R2 has serial interfaces in the network. Hrsp protocol is meant for ethernet interface or port. In fact if you try to configure hrsp under serial interface, you will get the message" unrecognised command"

From your diagram, it is obvious you are trying to provide gateway redundancy using hsrp, on 10.1.30/24 subnet which is connected to fe0/0 of Red router and fe/0 of router r1.

"The router (red) and the one below will be connected to a patch panel. I'm not sure if that connection will be on the same pins."

Here i assume " by same pins" you mean red router and the one below(r1) will be connected to same subnet(10.10.25.0/24). Even if they are not on the same pins( i.e different subnet) , it won't effect the rip updates.

However make sure you advertise the subnet 10.1.30.x/24 in rip or you will have reachability issue.

------------------------------------------

But you have problem in hsrp setup.

First of all r2 can not be a standby hsrp router for the reason mentioned above.

2nd thing, only red router is hsrp gateway for the host on the subnet 10.1.30.0/24.

You don't have redundancy for gateway.If your red router goes down, the host on 10.1.30.0/24 will be cut off from rest of network.

In order to provide gateway redundancy and to make the most of HSRP, you should configure r1 as standby not r2.

-----------------------------------------

You could manipulate hsrp on these routers, red and r1 so that both links will be used at a time by hosts on 10.1.30.0/24 subnet.

If you need more help how to do it, just ask

Thanks a lot!

franktuck Tue, 01/27/2009 - 05:45

Hi Sarah,

I hope I didn't add any confusion with the diagram. If you remove all the items in red that would be my current configuration. The 10.10.25.201 (P3) device is connected to the fe0/1 of the black router.

The red items depicts the proposed configuration.

Going forward both routers will have their s0 interface (10.10.65/24)connection to the same demarcation point.

HSRP will only be configured on the fe/0(10.1.30./24) of each router and each router track their own s0 interface.

Q: Will both routers receive the rip updates & user data?

Q: Can the 2811 support more that on default gateway?

sarahr202 Thu, 01/29/2009 - 06:31

Hi Frank!

"Going forward both routers will have their s0 interface (10.10.65/24)connection to the same demarcation point"

I will call red router as rR.

Now rR and R1 are connected by serial interface s0(10 .10.65.24).

How would these two routers r1 and rR be connected to r3?

Remember, you can connect single router on serial interface (if your are not using sub interface).

"HSRP will only be configured on the fe/0(10.1.30./24) of each router and each router track their own s0 interface."

Yes it will be fine , hosts on 10.1.30.0/24 can benefit from hsrp now. hsrp will not effect your rip updates and they both will receive user data as well if your are using two hsrp group , for one group say hsrp 1, r1 is active while rR is standby, for other group say hsrp 2 rR is actice while r1 is standby.

But will these two r1 and rR be able to forward ip packets from the hosts or receive the ip packets destined to host on 10.1.30.0/24?

The answer is if you connect these r1 and rR to properly to r3,and make sure you advertise the 10.1.30.0/24 subnet in rip so others routers will know about 10.1.30.0/24 subnet and thus be able to forward the packets destined to 10.1.30.0/24.

Q: Can the 2811 support more that on default gateway?

Yes,

If you need more help, just ask

franktuck Thu, 01/29/2009 - 20:00

Hi Sarah,

How would these two routers r1 and rR be connected to r3?

The routers rR & R1 will be connected to r3 via the channelbanks (microwave link)

Will I have to use two hsrp groups or was this just an example use in your response?

Thanks.

sarahr202 Fri, 01/30/2009 - 13:19

yes you have to use two hsrp group so both routers r1 and Rr will load balance the traffic and as standby router for each other.

But you can not connect r1 and Rr and r3 using micro wave link by serial ports unless you are using sub interface which you are not.

without sun interface, router expects only router on the serial link.

Thanks a lot!

Actions

This Discussion