Fast Ethernet Interface being take down with burst of traffic

[ciscobigcat]

Hi guys I have a problem which is really giving me a headache. It is a single router with dual ISPs. It is a 2800 and there is failover configured. I have implemented object tracking and the feature works great except that lately, whenever there is a lot of traffic coming perhaps from the internal users, we start getting intermittent outages.

I have gone deep into looking into this problem and have determined that our ISP#1 does not have any problems. What I think is happening is that whenever the router receives a lot of packets (30-40 users on the internal network) destined to the outside, the router CPU maybe gets too busy and the router then believes that the objects are no longer reachable and it triggers a failover which causes the router to re-direct traffic to the ISP#2. Then, because these are just quick burst of traffic, in the next 30 to 45 seconds after the router re-directed the traffic to ISP#2, the router object tracking engine detects that the objects are now again reachable and this then causes the router to re-direct all traffic back to ISP#1.

This cycle then continues all day and I need some help on how to prevent this.

Is there a way to perhaps tell the router to completely shut down the interface facing ISP#1 whenever there is a hicup and to keep it shutdown for at least 8 hours? This way I can prevent the router from going crazy.

Or is there a way to perhaps prevent this at the internal (LAN)  interface? Are there some metrics like QoS that I can implement on the internal Fast Ethernet Interface to prevent the burst of traffic from eating up the CPU?

thank you

[paolo bevilacqua]

Send "show interfaces" for the internet interfaces.

[ciscobigcat]

I checked on those and looks normal...

FastEthernet0/0/3 is up, line protocol is up
  Hardware is Fast Ethernet, address is 001e.xxx.xxxx (bia xxxx.xxxx.xxxx)
  MTU 1500 bytes, BW 100000 Kbit/sec, DLY 100 usec,
     reliability 255/255, txload 1/255, rxload 5/255
  Encapsulation ARPA, loopback not set
  Keepalive set (10 sec)
  Full-duplex, 100Mb/s
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:08, output never, output hang never
  Last clearing of "show interface" counters 1w2d
  Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue: 0/40 (size/max)
  5 minute input rate 2143000 bits/sec, 249 packets/sec
  5 minute output rate 206000 bits/sec, 172 packets/sec
     76583222 packets input, 2557951993 bytes, 0 no buffer
     Received 1330668 broadcasts, 0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 input packets with dribble condition detected
     96510627 packets output, 3055485812 bytes, 0 underruns
     0 output errors, 0 collisions, 0 interface resets
     26508 unknown protocol drops
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier
     0 output buffer failures, 0 output buffers swapped out