cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
738
Views
0
Helpful
0
Replies

HSRP default-gateway

Dam_Karp
Level 1
Level 1

Hi, 

 

I trust you are well.

 

Please find the attached network Topology created in GNS3.

 

I have a general inquiry on default-gateways. I set up HSRP between router 1 on fa0/0 (10.1.10.10) and router 2 on fa0/0 (10.1.10.11) to respond for 10.1.10.1 (Phantom IP/Virtual IP). This seems to be fine, because I can assign the default-gateway as 10.1.10.1 to all the host machines. 

 

Router 1 has the higher priority and preemption is enabled, therefore it will always be the active router. When I do a traceroute to the loopback interface on the server, it goes through the physical interface of router 1. This makes sense as I'm only running 1 standby group with router 2 being in the standby state.

 

The problem:

I ran two scenarios, both concerning the server.

#1: I had two static routes on the server to reach the 10.1.10.0/24 network, one path going through R1 (10.1.20.10) and the other through R2 (10.1.20.11). This allowed me to do load-balancing.

 

When R1 is still up and running and I ping the server (8.8.8.8) from the following hosts:

PC1: The server replies through router 1 (All successful).

PC2: The server replies through router 2 (All successful).

PC3: The server replies through router 1 (All successful).

 

When R1 is down, and R2 took over as the active router and I ping the server (8.8.8.8):

PC1: The server replies through router 1 (fail).

PC2: The server replies through router 2 (All successful).

PC3: The server replies through router 1 (fail).

 

#2: I changed some settings on the server: 

'no ip routing'

'ip default-gateway 10.1.10.1' (The Virtual/Phantom IP address)

All my pings are successful when from all hosts and all the Ping replies goes through R1. However, when R1 goes down (shutdown fa0/0), then all the pings fail. When I examine layer 2 on Wireshark, then all the replies are headed for the MAC address of R1, meaning it sends it's ICMP replies to R1 (Note: Only the fa0/0 is down and not fa0/1). All my pings heading for the interface Fa0/1 are successful (10.1.20.10 from Server). The replies are processed, yet they get dropped by the fa0/0 interface, which is shut down.

 

Note: I have static routes (on R1 and R2) pointing to the server's IP address (10.1.20.100) to reach the loopback adapter. Also, the Server is actually just another Cisco router.

 

Questions:

#1:

1. Does equal static routes use Round-Robin as the default-method for load-balancing and can it be changed?

2. Why does router 2 forward the replies from the server if it is the standby router? Shouldn't it drop the packets?

 

#2

3.1 What is best practice for the server? Meaning, if I use it with 'ip routing' enabled, what route should it use to get to the 10.1.10.0 network?

3.2 If I then decided to use it as a host, what should the default-gateway be?

 

4. Suppose the server is actually the ISP's router. How would the wiring from the two (1)routers->(to 2)demarcation zone->(to 3)ISP's router look (e.g. would it use two cables or only one)? Also, do they use the same Global public IP address for both routers the internal routers?

 

I think I might be missing something obvious. Please correct me, thank you.

Dam_Karp.

0 Replies 0