MLPPP traffic utilization

paul amaral · ‎01-27-2011

I hoping someone can give me some advise. I have a customer who is doing MLPPP on a cisco 1700 this terminates on a cisco 7200 on a CT3 single wide port card doing distributed multilink. The customer was complaining of "sluggish" performance and when i checked the routers for traffic utilization one T1 was hitting the max of 1.4 - 1.5Mb and the other around 900Kb to 1Mb while the multi link interface was around 2.4Mb. I know that multilink load balance will never be a perfect 50/50 but should one link being maxed out while the other is at 900Kb cause any issues?

I also looked at the traffic in a chart format and noticed alot of spikes, probabably due to global sync, whould i enabled WRED, if so do i need to enabled it on the physical interfaces or just on the logical MLPPP interface?

Another thing that I was not sure about is that i currently have fragmentation turned off but i was wondering if its recommend that i turn it on with interleaving? should I also change my MTU to something higher?

I still have to look at the router's logs to see if the problem is due to a physical link bouncing, taking errors, high CPU but any advise is apriciated.

thanks, Paul

1700 config:

interface Multilink1
description Bonded T1 interface.
bandwidth 3072
ip unnumbered Loopback1
ppp multilink
ppp multilink fragment disable
ppp multilink group 1
!
interface FastEthernet0
ip address xxxxx
speed auto
!
interface Serial0
description 2nd t1 - bonded t1
bandwidth 1536
no ip address
encapsulation ppp
no fair-queue
ppp multilink
ppp multilink group 1
!
interface Serial1
description
bandwidth 1536
no ip address
encapsulation ppp
ppp multilink
ppp multilink group 1

Multilink1, bundle
Bundle up for 2w1d, total bandwidth 3072, load 9/255
Receive buffer limit 24384 bytes, frag timeout 1000 ms
    0/0 fragments/bytes in reassembly list
    925 lost fragments, 23799 reordered
    4768/5553936 discarded fragments/bytes, 4768 lost received
    0xCF451C received sequence, 0x3509C9 sent sequence
Member links: 2 active, 0 inactive (max not set, min not set)
    Se0, since 2w1d
    Se1, since 2w1d

7200:

Multilink1, bundle name is woods-hole-gw
Bundle up for 2w1d
Bundle is Distributed
0 lost fragments, 8526621 reordered, 0 unassigned
0 discarded, 0 lost received, 117/255 load
0x331A58 received sequence, 0xCC9C28 sent sequence
Member links: 2 active, 0 inactive (max not set, min not set)
Se10/1/0/19:1, since 2w1d, no frags rcvd
Se10/1/0/17:1, since 2w1d, no frags rcvd

paul amaral · ‎01-27-2011

i just noticed on the 1700 side that there seems to be alot lost/reorder frames.

show ppp multilink

Multilink1, bundle name is core.prv.ri
Bundle up for 00:05:25, total bandwidth 3072, load 37/255
Receive buffer limit 24384 bytes, frag timeout 1000 ms
    1/1430 fragments/bytes in reassembly list
    646 lost fragments, 12877 reordered
    3229/4027791 discarded fragments/bytes, 3229 lost received
    0xC71B received sequence, 0xAEC9 sent sequence
Member links: 2 active, 0 inactive (max not set, min not set)
    Se1, since 00:05:25
    Se0, since 00:05:25

and seeing some input errors on the multilink interface

Multilink1 is up, line protocol is up
Hardware is multilink group interface
Description: Bonded T1 interface.
Interface is unnumbered. Using address of Loopback1 (xxxx
MTU 1500 bytes, BW 3072 Kbit, DLY 100000 usec,
     reliability 255/255, txload 62/255, rxload 78/255
Encapsulation PPP, LCP Open, multilink Open
Listen: CDPCP
Open: IPCP, loopback not set
Keepalive set (10 sec)
DTR is pulsed for 2 seconds on reset
Last input 00:00:00, output never, output hang never
Last clearing of "show interface" counters 00:10:54
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 942000 bits/sec, 168 packets/sec
5 minute output rate 756000 bits/sec, 184 packets/sec
     117547 packets input, 100185035 bytes, 0 no buffer
     Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
     8763 input errors, 0 CRC, 7319 frame, 122 overrun, 0 ignored, 0 abort
     117730 packets output, 54529449 bytes, 0 underruns
     0 output errors, 0 collisions, 0 interface resets
     0 output buffer failures, 0 output buffers swapped out
     0 carrier transitions

however the serial T1 interfaces have no input errors.

cisco 1721 (MPC860P) processor (revision 0x400) with 54534K/11002K bytes of memory.
Processor board ID FOC09021CQN (2749784902), with hardware revision 0000
MPC860P processor: part number 5, mask 2
Bridging software.
X.25 software, Version 3.0.0.
1 FastEthernet/IEEE 802.3 interface(s)
2 Serial network interface(s)
WIC T1-DSU
32K bytes of non-volatile configuration memory.
32768K bytes of processor board System flash (Read/Write)

paul amaral · ‎03-15-2011

this was fixed by allowing fragmentation by removing ppp multilink fragment disable. i also enabled interleaving and the performance was huge.

Richard Burts · ‎03-15-2011

Paul

Thanks for posting back to the forum indicating that the problem was solved, what the problem seemed to be, and what you did to resolve the problem. It makes the forum more useful when people can read about an issue and can read how that issue was resolved. Your problem was an interesting one and may be instructive to readers of the forum.

HTH

Rick

HTH

Rick

paul amaral · ‎03-16-2011

no problem, its why we are here to learn and hopefully share.