12-28-2009 11:59 AM - edited 03-06-2019 09:05 AM
Dear Concern,
I have two sites and configured Layer 3 ether channel. But facing a problem like that when microwave link goes down site A is disconnected for 1.5 min and after 1.5 min it comes up. But when fiber link goes down then site A does not disconnected, it works normally.
Please find attachment for configuration and diagram. Urgent issue
Thanks in advance.
12-28-2009 12:06 PM
Hello,
given the different nature of the two links, I would use each of them as an indipendent layer 3 link with a routing protocol running over both.
in this way the routing protocol can detect the failure of a link and can adapt to this by using only the other link.
for example you could use EIGRP or OSPF
Hope to help
Giuseppe
12-28-2009 12:25 PM
Thanks for your quick response. What do u mean by nature of the links ? please clarify.
Can you help me about pagp times ? how can I define it ?
12-28-2009 12:59 PM
Hello,
I try to repeat I would use the two links as two different L3 links
example
default interface gi1/47
default interface gi1/48
int gi1/47
no switchport
ip address 192.168.x1.y1 255.255.255.252
int gi1/48
no switchport
ip address 192.168.x2.y2 255.255.255.252
router eigrp100
network 192.168.x1.0
network 192.168.x2.0
no auto-summary
! other network commands for nets to be advertised
note: the other end must use the same EIGRP AS number (100 in my example)
if one link is fiiber and the other one is based on a RF microwave link they have different nature.
For this reason I would use each with its own IP subnet.
With default timers EIGRP detects a failure in 15 seconds
Hope to help
Giuseppe
12-30-2009 02:23 AM
Thanks for the update. I have implemented OSPF and it takes 30 seconds for link failure issue. Since OSPF is running in existing system thats why client dont agree to use EIGRP. Actually huge voice and data traffic will be passed through the links so there is absolately no tollarance of delay , it should be zero delay.
Is there any way to overcome this delay issue ?
Please update .
thanks a lot !
12-30-2009 03:19 AM
Hello Skmdimram,
I've used EIGRP just to make an example.
With default settings OSPF can take up to 40 seconds to detect a dead neighbor.
Depending on what your devices support you can use:
sub-second hello interval
see in interface config mode you can use:
ip ospf dead-interval {seconds | minimal hello-multiplier multiplier}
http://www.cisco.com/en/US/docs/ios/iproute/command/reference/irp_osp1.html#wp1014268
minimal means 1 second of dead interval.
better results with BFD bidirectional forwarding detection that allows to implement IP fast convergence. BFD is a lightweight protocol that checks ip connectivity and provides its services to routing protocols including OSPF.
You can enable BFD for a subset of the interfaces for which OSPF is routing by using the ip ospf bfd command in interface configuration mode.
see
you need to verify what features are supported on your devices. if they are multilayer switches and they are not C6500 I'm afraid BFD is not supported.
You can use www.cisco.com/go/fn
to check with feature navigator
I've tried and I get only C3750 Metro ethernet or ME 3400 among multilayer switches
Hope to help
Giuseppe
01-02-2010 11:24 AM
Thanks a lot for sharing a nice document about BFD, its really help me .
Would you please suggest me how I will find out that OSPF load balancing are happening between two links ? Actually I want to know , is there any command that will show load balanced traffic .
Many thanks.
Regards
Imran
01-02-2010 11:30 AM
Hello Imram,
sh ip cef exact-route
can tell you on what link given a source and destination address of the flow the packet is sent.
Other more platform specific CEF show commands may exist for your devices.
example
sh ip cef exact-route 1.1.1.1 ?
A.B.C.D Destination address of session
BFD should be fine with your 7609 devices.
Hope to help
Giuseppe
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: