EIGRP what causes excessive RTO

[jim-bell]

What causes excessive RTO times between EIGRP neighbors.

We have a transit vlan40 connecting 3 wan routers (7206's ) and 2 MSFC routers all connectted over a switched infrastructure (on vlan40) see output below;

The 10.20.40.4 address is the other MSFC02, why is this RTO (of 3000) so high in comparison to the others (200)

MSFC01#sh ip eigrp neigh

IP-EIGRP neighbors for process 1

H Address Interface Hold Uptime SRTT RTO Q Seq Type

(sec) (ms) Cnt Num

3 10.20.40.4 Vl40 14 01:01:00 1 3000 0 136

2 10.20.40.5 Vl40 12 01:12:40 8 200 0 186

1 10.20.40.1 Vl40 13 01:12:40 7 200 0 272

0 10.20.40.2 Vl40 14 01:12:40 3 200 0 650

MSFC01#sh ip eigrp neigh det

IP-EIGRP neighbors for process 1

H Address Interface Hold Uptime SRTT RTO Q Seq Type

(sec) (ms) Cnt Num

3 10.60.40.4 Vl50 12 01:03:51 1 3000 0 136

Version 12.1/1.2, Retrans: 1, Retries: 0

2 10.20.40.5 Vl50 11 01:15:30 8 200 0 186

Version 12.2/1.2, Retrans: 0, Retries: 0

1 10.20.40.1 Vl50 11 01:15:30 7 200 0 272

Version 12.1/1.2, Retrans: 0, Retries: 0

0 10.20.40.2 Vl50 14 01:15:30 2 200 0 659

Version 12.1/1.2, Retrans: 0, Retries: 0

The MSFC's are on 2 x 6500 series switches which are trunked together, with 2 wan routers off one 6500 switch and the other off the 2nd 6500 switch.

[#2]

SRTT(Smooth round-trip time)is the number of milliseconds it takes for an EIGRP packet to besent to this neighbor and for the local router to receive an acknowledgment of that packet.

RTO(Retransmission timeout in milliseconds)is the amount of time the software waits before retransmitting a packet from the retransmission queue to a neighbor.

The SRTT timer can be used as an indication about the round trip delay on a particular link. It is used to calculate the RTO value.

So, normally if you have a high RTO, it is because you have a high SRTT. A high SRTT value indicates the long amount of time that passed before the router received an acknowledgment from its neighbor. Possible causes could be that the interface or the router's CPU is taking a heavy load. But here in your case, since the SRTT value is not high,this is not the case.RTO changes everytime another packet has to be retransmitted. RTO is initially set to a value that is computed by TCP from the SRTT, current pacing timer or 200 (the highest of those 3 values). The value of RTO is not necessarily used to detemine what is going on for the interface or if there is a potential problem concerning EIGRP. It is merely a timer that is used to determine how long the IOS waits before retransmitting a packet to a neighbor that failed to acknowledge a packet.

[jim-bell]

Thanks for your reply, that has helped.

The RTO timer did eventually came down to the same as the other neighbors around 200, after a period of around 15 hours.