High errors on multilink interface but no or very less errors on physical serial interfaces

[saurab mattoo]

I have 4 T1's & the 4 physical interfaces are showing negligible errors but the virtual multilink interface is showing high number of errors. No issues with reliability, Tx/ Rx load & no carrier transitions(on physical & virtual interface); only the circuit is over utilized which I believe should not cause input/ CRC errors to generate. What could be the possible issue for this behaviour e.g.

1. Cisco bug? It is a cisco 2811 having 12.4 IOS version.

OR/ AND

2. Will rebuilding the multilink interface resolve the issue.

OR/ AND

3. Should the cables on the 4 serial interfaces be changed.

Please let know your comments!

[Georg Pauwen]

Hello Saurab,

this could be due to the bug below:

/sh int/ on MLP Bundle interface shows 100% input errors

CSCsj26883

Description

Symptom: on a multilink bundle interface, the output from "show interface"
will show an input error count which includes essentially 100% of the traffic
being received. Other drop counters, usually the "bad frame" counter,
are also incremented to match. For example:

# show interface Multilink1
Multilink1 is up, line protocol is up
Hardware is multilink group interface
Internet address is 10.20.0.1/30
MTU 1500 bytes, BW 3072 Kbit, DLY 100000 usec,
reliability 181/255, txload 1/255, rxload 1/255
Encapsulation PPP, LCP Open, multilink Open
...
5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 0 bits/sec, 0 packets/sec
6080 packets input, 1027997 bytes, 0 no buffer
Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
6077 input errors, 0 CRC, 6077 frame, 0 overrun, 0 ignored, 0 abort
^^^^ ^^^^

Impact: Very Low. This is an accounting error; the counters are
being accidentally incremented each time Multilink begins to receive
a new datagram. This does not affect actual traffic flow or operation
of the bundle, it is only the statistics that are in error.

Workaround: none.

[saurab mattoo]

Thanks gpauwen,

The router is having 365000 input errors for received 123 milllion packets. Could it be related to this bug.

[Georg Pauwen]

Hello Saurab,

clear the counters for the interface and check if there is a linear increase for input errors and CRC errors. That is what the bug describes should be happening.

[saurab mattoo]

Ok, I will do that & observe for a linear increase.

Suppose it is related to this bug; what is the workaround; as per CSCsj26883 it is showing as fixed. Does that mean it should not occur anymore or is some IOS change required.

[Georg Pauwen · Accepted solution]

Saurab,

IOS upgrade is recommended. The affected releases are:

12.4(15.6)
12.4(16)
12.4(16.9)

Any of the following releases contain the fix:

12.2(33)IRC
12.2(33)MRA
12.2(33)SB14
12.2(33)SB3
12.2(33)SB4
12.2(33)SB5
12.2(33)SB6a
12.2(33)SB6aa
12.2(33)SB6b
12.2(33)SB7
12.2(33)SB8
12.2(33)SB8b
12.2(33)SCB
12.2(33)SCB1
12.2(33)SCB3
12.2(33)SCB6
12.2(33)SCC
12.2(33)SCC1
12.2(33)SCC2
12.2(33)SCD
12.2(33)SCD1
12.2(33)SCF5
12.2(33)SCG2
12.2(33)SCG3
12.2(33)SCG4
12.2(33)SCG5
12.2(33)SCH
12.2(33)SCH1
12.2(33)SRC3
12.2(33)SRC4
12.2(33)SRC5
12.2(33)SRD
12.2(33)SRD1
12.2(33)SRD2
12.2(33)SRD3
12.2(33)SRD4
12.2(33)SRE
12.2(33)SRE5
12.2(33)SRE8
12.2(33)XND
12.2(33)XNE
12.2(33)XNE1
12.2(34)SB1
12.2(34)SB2
12.2(34)SB3
12.2(34)SB4
12.2(34)SB4a
12.2(34)SB4b
12.2(34)SB4c
12.3(23.9)
12.4(15)MD2
12.4(15)T10
12.4(15)T11
12.4(15)T12
12.4(15)T13
12.4(15)T14
12.4(15)T16
12.4(15)T17
12.4(15)T2
12.4(15)T6a
12.4(15)T8
12.4(15)XL3
12.4(15)XQ2a
12.4(15)XQ2b
12.4(15)XQ4
12.4(15)XR4
12.4(16.10)
12.4(16.10)T
12.4(16a)
12.4(18a)JA1
12.4(19)MR2
12.4(19)MR3
12.4(20)MR
12.4(20)T2
12.4(20)T3
12.4(20)T4
12.4(21a)JA
12.4(21a)JA1
12.4(21a)JA2
12.4(21a)JY
12.4(22)MD
12.4(22)MD1
12.4(22)MDA1
12.4(22)MDA2
12.4(22)MDA3
12.4(22)T1
12.4(22)T2
12.4(22)T3
12.4(22)T4
12.4(22)T5
12.4(22)XR3
12.4(22)YB
12.4(22)YB1
12.4(22)YB2
12.4(22)YB8
12.4(22)YD
12.4(22)YE2
12.4(22)YE4
12.4(22.3.4)PIC1
12.4(23)
12.4(23a)
12.4(23b)
12.4(23c)
12.4(23c)JA3
12.4(23c)JA7
12.4(24)GC1
12.4(24)MD
12.4(24)MDA13
12.4(24)MDB13
12.4(24)MDB14
12.4(24)MDB15
12.4(24)T
12.4(24)T1
12.4(24)T10
12.4(24)T2
12.4(24)T3
12.4(24)T4j
12.4(24)T4k
12.4(24)T4l
12.4(24)T4m
12.4(24)T4n
12.4(24)T5
12.4(24)T6
12.4(24)T7
12.4(24)T8
12.4(24)T9
12.4(24)YE
12.4(24)YE3e
12.4(24)YG1
12.4(24)YG2
12.4(24)YS
12.4(24)YS1
12.4(24)YS2
12.4(24)YS3
12.4(24)YS4
12.4(24.5.2)PIC1
12.4(25a)
12.4(25c)
12.4(25d)
12.4(25d)JA1
12.4(25e)JAL1
12.4(25e)JAL1a
12.4(25e)JAL2
12.4(25e)JAM
12.4(25e)JAM2
12.4(25e)JAN1
12.4(25e)JAZ
12.4(25f)
12.4(25g)
15.0(1)M
15.0(1)M1
15.0(1)M10
15.0(1)M2
15.0(1)M4
15.0(1)M7
15.0(1)M8
15.0(1)M9
15.0(1)XA
15.0(1)XA1
15.1(1)MR6
15.1(1)T
15.1(1)T1
15.1(1)T3
15.1(1)T4
15.1(1)T5
15.1(1)XB1
15.1(2)T
15.1(2)T1
15.1(2)T2
15.1(2)T3
15.1(2)T4
15.1(2)T5
15.1(3)MRA
15.1(3)MRA1
15.1(3)S2
15.1(3)S5
15.1(3)S5a
15.1(3)T
15.1(3)T1
15.1(3)T2
15.1(3)T3
15.1(3)T4
15.1(4)GC1
15.1(4)M
15.1(4)M1
15.1(4)M2
15.1(4)M3
15.1(4)M4
15.1(4)M5
15.1(4)M6
15.2(1)S
15.2(1)S1
15.2(1)T
15.2(1)T1
15.2(1)T2
15.2(1)T4
15.2(2)JA
15.2(2)JA1
15.2(2)JAX
15.2(2)JB
15.2(2)JB2
15.2(2)SA
15.2(2)SA1
15.2(2)SA2
15.2(2)SNI
15.2(2)T
15.2(2)T1
15.2(2)T3
15.2(2)T4
15.2(3)GC
15.2(3)GC1
15.2(3)GCA
15.2(3)GCA1
15.2(3)T
15.2(3)T1
15.2(3)T2
15.2(3)T3
15.2(4)JA
15.2(4)JA1
15.2(4)M1
15.2(4)M2
15.2(4)M3
15.2(4)M4
15.2(4)S1c
15.2(4)S2
15.2(4)S3
15.2(4)S3a
15.2(4)S4
15.2(4)XB11
15.3(1)S1
15.3(1)S1e
15.3(1)S2
15.3(1)T
15.3(1)T1
15.3(1)T2
15.3(2)S
15.3(2)S1
15.3(2)S1b
15.3(2)S2
15.3(2)T
15.3(2)T1
15.3(3)M
15.3(3)S

[saurab mattoo]

Thanks for the support, I will update you on this.

[saurab mattoo]

Cisco IOS Software, 2800 Software (C2800NM-IPBASEK9-M), Version 15.1(3)T, RELEASE SOFTWARE (fc1)

I guess it is not related to this bug..The IOS is 15.1(3)T

[saurab mattoo]

I verified that the errors are incrementing linearly & the IOS is not among the reported ones. What do you suggest; replacing the cables on serial interface?

[saurab mattoo]

sorry, not increasing linearly.

[Georg Pauwen]

Saurab,

since you are not seeing any errors on the actual physical interfaces, and you are not experiencing any reliability issues, my guess is that this is simply a cosmetic error. Is this a production environment ? Obviously you don't want to see errors on your interfaces, I wouldn't either. If you have ample downtime available, changing the cables would be an option. If you decide to do that, I would be interested to know if the errors disappeared.

[saurab mattoo]

Yes, it is a production environment. I will let know the results once we are done with cable change activity.

[Georg Pauwen]

Saurab,

I have googled around, and apparently this same problem does also occur in later releases. In the post below, your problem is being observed, and the IOS is 15.0M(1). I am not sure if Cisco accurately updates the bug fix releases and affected releases...

https://supportforums.cisco.com/discussion/11950356/multilink-interface-taking-multiple-input-frame-and-overrun-errors

[saurab mattoo]

Danny has not mentioned what exactly resolved the issue :(

My next step is to replace the serial cables & if that too does not resolve then replace the router :)