IGP Cost

[ricaela]

Hi Guys,

1. Is there gonna be a problem if two router interfaces cost doesn't match? thanks!!

example ,

r1 g0/0 cost is 1-----------------r2 g0/0 cost is 10

if no issue, then total cost for the link is 11??

2. also what does this output mean? it is being load balanced right? can u further explain the output shown below:

R1#traceroute 4.4.4.4

Type escape sequence to abort.
Tracing the route to 4.4.4.4

1 10.0.12.2 32 msec
10.0.13.2 28 msec
10.0.12.2 32 msec
2 10.0.34.2 56 msec
10.0.24.2 64 msec
10.0.34.2 60 msec

refer to the attached photo for the topology .

thank you!!

3. for the topology attached,

when i check r1 ip route going to 10.0.34.2 , why is it not load balanced? (all cost are same)

O 10.0.34.0/29 [110/3] via 10.0.12.2, 00:14:04, GigabitEthernet0/0

why is it it is not like this:

O 10.0.34.0/29 [110/3] via 10.0.12.2, 00:14:04, GigabitEthernet0/0

                                     via 10.0.13.2,               , G1/0

THANK YOU!!!

[paul driver]

Hello
Manually setting the interface ospf cost is locally significant it allows the router to calculate the preferred path as such neighbour(s) on the other side of the connection with different ospf interface cost value isn’t applicable to the local rtrr but i would say they should have parity on either side.

Also ospf by default performs ecmp on all it links however manual intervention such as changing the ospf reference bandwidth or interface cost can negate this and even cause suboptimal routing.

[Joseph W. Doherty]

"1. Is there gonna be a problem if two router interfaces cost doesn't match?"

Depends on what you're trying to accomplish (also the nature of your links - certain technologies actually support different [available] ingress vs. egress bandwidths).

"if no issue, then total cost for the link is 11??"

Generally no.  Cost is usually calculated just for egress (which is the case for OSPF).

"2. also what does this output mean? it is being load balanced right? can u further explain the output shown below:"

Load balanced? Depends how you define LB.  OSPF will round-robin flows across equal cost paths, to destination, but does not account for actual link loading.

"3. for the topology attached,

when i check r1 ip route going to 10.0.34.2 , why is it not load balanced? (all cost are same)"

Even if link costs are the same, 10.0.34.0/29 is "closer" to R1 via R3, whereas the longer path is R2 and R4.

[Martin L]

No issues, R1 cost to reach R2 is 1 while R2 cost to R1 is 10. 

no, total cost for the link is 1 (or 10 via other way); you may see 11 (10 +1) if you ping Loopback interface or ping from Loopback.

OSPF does equal load balance by default so you have 2 paths from R1 to R4, R2 to R3 but only 1 "best" path R1 to R2; you may see R1 can reach R2 via R3,4 when R1 to R2 link fails.

if link is not load balanced, double check costs; record cost of active link, then shutdown active path and compare costs.

Regards, ML
**Please Rate All Helpful Responses **