PRI backup drops packets

[simonthompson]

I have 1 PRI backing up 2 by 2 meg IPLs on seperate routers. When both IPLs fail and the PRI kicks in, I lose packets on a simple ping over the PRI.

I have been told that it is due to the additional processing required by the router to split the packets up for the ppp connections and then re-assembling them at the other end of the PRI. Is this correct, and if so, is there a solution?

Thanks.

[lgijssel]

This does not directly sound like a viable explanation to me.

There might be an issue with CPU load but then you should have the CPU running at 50% or more all the time. You can check this through the command: sh proc

First I would advise you to check that your router meets the memory-requirements for your current release and to check the config. You may post it if you cannot manage this yourself.

You should also check the load on the PPP-link that loses packets. If that link is at a high load, this is probably your explanation. However, if everything else is ok, you should have no problems during non-office hours.

Regards,

Leo

[simonthompson]

The router has 128 meg of memory. 5 minute CPU average is 7%.

We ran a test last night on it and there were a lot of dropped packets on the input queue as in the sh int sum below:

Interface IHQ IQD OHQ OQD RXBS RXPS TXBS TXPS TRTL

---------------------------------------------------------------------

Serial2/0:2 0 531463 0 0 0 0 0 0 0

Serial2/0:3 0 531389 0 0 0 0 0 0 0

Serial2/0:4 0 535183 0 0 0 0 0 0 0

Serial2/0:5 0 535279 0 0 0 0 0 0 0

Serial2/0:6 0 543203 0 0 0 0 0 0 0

Serial2/0:7 0 543379 0 0 0 0 0 0 0

Serial2/0:8 0 539438 0 0 0 0 0 0 0

Serial2/0:9 0 541761 0 0 0 0 0 0 0

Serial2/0:10 0 543281 0 0 0 0 0 0 0

Serial2/0:11 0 543494 0 0 0 0 0 0 0

Serial2/0:12 0 540644 0 0 0 0 0 0 0

Serial2/0:13 0 565868 0 0 0 0 0 0 0

Serial2/0:14 0 465781 0 0 0 0 0 0 0

* Serial2/0:15 0 360719 0 0 0 0 0 0 0

Serial2/0:16 0 522166 0 0 0 0 0 0 0

Serial2/0:17 0 544311 0 0 0 0 0 0 0

Serial2/0:18 0 544603 0 0 0 0 0 0 0

Serial2/0:19 0 546292 0 0 0 0 0 0 0

Serial2/0:20 0 543014 0 0 0 0 0 0 0

Serial2/0:21 0 569401 0 0 0 0 0 0 0

Serial2/0:22 0 462669 0 0 0 0 0 0 0

Serial2/0:23 0 14 0 0 0 0 0 0 0

Serial2/0:24 0 29 0 0 0 0 0 0 0

Serial2/0:25 0 18 0 0 0 0 0 0 0

Serial2/0:26 0 25 0 0 0 0 0 0 0

Serial2/0:27 0 59 0 0 0 0 0 0 0

Serial2/0:28 0 3708 0 0 0 0 0 0 0

Serial2/0:29 0 46 0 0 0 0 0 0 0

Serial2/0:30 0 80 0 0 0 0 0 0 0

[tepatel]

Sounds link you are getting packet drops on multilink ppp links on pri during backup. Now that packet loos could be because of many issues. We need to verify

1)bundeling is occured properly

2)MRRU negotiations for multilink

3)route-cache etc.

To troubleshoot that issue here are some of the links which will help

For "Troubleshooting Input Queue Drops and Output Queue Drops"

http://www.cisco.com/warp/public/63/queue_drops.html

Also following link for async mlppp troubleshooting have step by step assistance based on a problem.

http://www.cisco.com/univercd/cc/td/doc/cisintwk/intsolns/asynmppp/mppptr.htm

Try to verify "poor performance" section.