Redudancy with IP sla not work as expected

[nicanor00]

Hi

I have configured Reliable Static Routing Backup Using Object Tracking on the router at the remote site

But it is not work as i want (found architecture in attachement

I need the trafig going from the remote site to be send to the HEADQUARTER, then when link to the HEADQUARTER fail, router check the link for 1 min then after 1 min if the link remain down, trafic is send to the BACKUP site

Router continu check HEADQUARTER link and as soon as Headquarter link is up during 1 min, trafic is switched and send on HEADQUARTER link

But it is not work

Please helps

[Philip D'Ath]

Basic jist looks correct to me.  What is not working?  What is not happening that you would like to happen?

[nicanor00]

Hi Thanks for your answer In the show run that I send, I use server IP (192.168.1.5) to test, it doesnt work ip sla 10 icmp-echo 192.168.1.5 source-interface FastEthernet0/1 timeout 1000 threshold 5 frequency 10 ip sla schedule 10 life forever start-time now The only route that is use is backup route when I shutdown F0/0 : 192.168.4.2 interface on backup router, trafic is not redirected on the Headquarter But I use F0/0: 192.168.5.2 it work fine ip sla 10 icmp-echo 192.168.5.2 source-interface FastEthernet0/1 timeout 1000 threshold 5 frequency 10 ip sla schedule 10 life forever start-time now Trafic is redirected on backup site when I shutdown F0/0: 192.168.5.2 And the trafic going on headquarter when I do no shut on F0/0: 192.168.5.2 Why It cannot work when I use this ip sla 10 icmp-echo 192.168.1.5 source-interface FastEthernet0/1 timeout 1000 threshold 5 frequency 10 ip sla schedule 10 life forever start-time now Regards

[Philip D'Ath]

When it is in its normal state can you note the output of "show ip sla statistics 10" and "show track 10".  Then shutdown Fa0/0 at HW, wait a minute, and note the output of the same two commands again.  Then post both outputs here please.

[nicanor00]

Hi

Kindly found the output of each test in attachement

Regards

[Philip D'Ath]

ip sla itself uses the "routing rules" the router is using for sending test packets.

Everything works perfectly when you use 192.168.5.2 as a test target, because that IP address is on a directly connected interface.  If the interface/circuit goes down the router is invalid, and if the interface comes back up the route becomes valid again.

However pinging the server 192.168.1.5 is not the same.

When the circuit is up, you have a default route going via the 192.168.5.0/30 path, and the ping works.  When the circuit goes down ip sla fails, and the default route now points to 192.168.4.0/30.

Now this is the catch - when your circuit comes back up your pings will obey the routing table, and keep going down the 192.168.4.0/30 path, which will never work, and hence it will also show as failed and never come back up.

The easiest fix is to put a static permanent route for 192.168.1.5 via 192.168.5.2, so the IP sla traffic for that server always goes via one path, even if it is down.  The other option is if 192.168.1.0/24 is only reachable via 192.168.5.2 then add a permanent static route for the whole subnet.