| Attachment | Size |
|---|---|
 10K.txt | 60.53 KB |
Got a bit of an odd one here. Have been using 16 channels using a Cisco Ubr10k w/ PRE5s (SCJ4) and 20x20 cards and there seems to be a large amount of packet loss if all 16 channels are broadcast to a node.
The odd part is, if I shutdown 12 of the 16 channels and only have 4, then the packet loss virtually disappears. This happens if I use any set of the 4 channels through any port and on any card (4 cards). Same issue applies if I use 8 or 12 channels. I do not see any interference with the spectrum analyzer and have tried different frequencies - same issue every time. We are checking by pinging IPs from the CMTS and from the provisioing server. Sanitized config is attached, does anyone have suggestions as to what may be causing this?
Thanks in advance!
You could try to remove the upstream statement:
interface Cable7/0/0description Service Group 3
downstream Integrated-Cable 7/0/0 rf-channel 0-3
downstream Integrated-Cable 7/0/1 rf-channel 0-3
downstream Integrated-Cable 7/0/2 rf-channel 0-3
downstream Integrated-Cable 7/0/3 rf-channel 0-3
besides this, you should reduce the amount of primary channels, because they will eat up some bandwidth.
Dont tie the US channels to 7/0/0 DS 0-3
Unless you have mainly D2.0 modems, dont make every DS channel primary.
You can simplify your DS bonding groups
Add another US bonding group
Consider reducing your wiring by using channel stacking
Consider using 1 x 6.4 US rather than 2 x 3.2 US
Something like below
!
interface Cable7/0/0
description Service Group 3
!--- Change so only every 4th channel is primary
downstream Integrated-Cable 7/0/0 rf-channel 0
downstream Integrated-Cable 7/0/1 rf-channel 0
downstream Integrated-Cable 7/0/2 rf-channel 0
downstream Integrated-Cable 7/0/3 rf-channel 0
cable mtc-mode
no cable packet-cache
cable default-phy-burst 0
cable bundle 1
cable upstream max-ports 8
cable upstream bonding-group 700
upstream 0
upstream 1
upstream 2
upstream 3
upstream 4
upstream 5
upstream 6
upstream 7
attributes 80000000
cable upstream bonding-group 701
upstream 0
upstream 1
upstream 2
upstream 3
attributes 80000000
!--- add this
cable upstream bonding-group 702
upstream 4
upstream 5
upstream 6
upstream 7
attributes 80000000
cable upstream 0 connector 0
cable upstream 0 frequency 19600000
cable upstream 0 channel-width 3200000 3200000
cable upstream 0 ingress-noise-cancellation 200
cable upstream 0 docsis-mode atdma
cable upstream 0 minislot-size 2
cable upstream 0 range-backoff 3 6
cable upstream 0 modulation-profile 221
cable upstream 0 equalization-coefficient
no cable upstream 0 shutdown
!--- consider using channel stack
cable upstream 1 connector 0
cable upstream 1 frequency 22800000
cable upstream 1 channel-width 3200000 3200000
cable upstream 1 ingress-noise-cancellation 200
cable upstream 1 docsis-mode atdma
cable upstream 1 minislot-size 2
cable upstream 1 range-backoff 3 6
cable upstream 1 modulation-profile 221
cable upstream 1 equalization-coefficient
no cable upstream 1 shutdown
cable upstream 2 connector 2
cable upstream 2 frequency 26000000
cable upstream 2 channel-width 3200000 3200000
cable upstream 2 ingress-noise-cancellation 200
cable upstream 2 docsis-mode atdma
cable upstream 2 minislot-size 2
cable upstream 2 range-backoff 3 6
cable upstream 2 modulation-profile 221
cable upstream 2 equalization-coefficient
no cable upstream 2 shutdown
!--- consider using channel stack
cable upstream 3 connector 2
cable upstream 3 frequency 29200000
cable upstream 3 channel-width 3200000 3200000
cable upstream 3 ingress-noise-cancellation 200
cable upstream 3 docsis-mode atdma
cable upstream 3 minislot-size 2
cable upstream 3 range-backoff 3 6
cable upstream 3 modulation-profile 221
cable upstream 3 equalization-coefficient
no cable upstream 3 shutdown
cable upstream 4 connector 4
cable upstream 4 frequency 32400000
cable upstream 4 channel-width 3200000 3200000
cable upstream 4 docsis-mode atdma
cable upstream 4 minislot-size 2
cable upstream 4 range-backoff 3 6
cable upstream 4 modulation-profile 221
cable upstream 4 equalization-coefficient
no cable upstream 4 shutdown
!--- consider using channel stack
cable upstream 5 connector 4
cable upstream 5 frequency 35600000
cable upstream 5 channel-width 3200000 3200000
cable upstream 5 docsis-mode atdma
cable upstream 5 minislot-size 2
cable upstream 5 range-backoff 3 6
cable upstream 5 modulation-profile 221
cable upstream 5 equalization-coefficient
no cable upstream 5 shutdown
cable upstream 6 connector 6
cable upstream 6 frequency 38800000
cable upstream 6 channel-width 3200000 3200000
cable upstream 6 docsis-mode atdma
cable upstream 6 minislot-size 2
cable upstream 6 range-backoff 3 6
cable upstream 6 modulation-profile 221
cable upstream 6 equalization-coefficient
no cable upstream 6 shutdown
!--- consider using channel stack
cable upstream 7 connector 6
cable upstream 7 frequency 16400000
cable upstream 7 channel-width 3200000 3200000
cable upstream 7 docsis-mode atdma
cable upstream 7 minislot-size 2
cable upstream 7 range-backoff 3 6
cable upstream 7 modulation-profile 221
cable upstream 7 equalization-coefficient
no cable upstream 7 shutdown
interface Wideband-Cable7/0/0:0
cable bundle 1
cable rf-channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel 3 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 3 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 3 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 3 bandwidth-percent 5 remaining ratio 100
!
interface Wideband-Cable7/0/0:1
cable bundle 1
cable rf-channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel 3 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 1 channel 3 bandwidth-percent 5 remaining ratio 100
!
interface Wideband-Cable7/0/0:2
cable bundle 1
cable rf-channel controller 2 channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 2 channel 3 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 0 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 1 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 2 bandwidth-percent 5 remaining ratio 100
cable rf-channel controller 3 channel 3 bandwidth-percent 5 remaining ratio 100
!
Thanks for the suggestions, removing the upstream statement and primary channels seems to be helping a bit.
So a 16x4 modem would perform better with a 16x2 bonding group using 2x 6.4 width rather than 4x 3.2 width (assuming US SNR is solid at 36.1)?
1 x 6.4 is more efficient than 2 x 3.2
You do lose 3dB SNR though with the wider channel.
So if there is noise you are best to configure auto-downgrade 6.4 3.2
Or on low frequencies you may be best to stick with 3.2