| Attachment | Size |
|---|---|
 BSR 1000 US03 SNR historical | 4.93 KB |
Hi group
I'm having a trouble with a bsr1000. All the upstreams port from the cmts goes from SNR 0 to 280 and viceversa, and the upstream port 3 remains on SNR 0.
We have another cmts, a bsr2000 and we do not have this issues. What I tried was to set another channel-width and frequency but no result.
I don't think this could be noise because it's in all upstream port (except No. 3 which remains SNR 0). In logs I don't see any strange but next:
07/15-02:19:10.02:BCM]-W- BCM.6 Dropped Ds Msg Pkt: mdrvSendMsgDs() called while msg wait queue is full, restarting 3210
[07/15-02:19:10.99:BCM]-W- BCM.7 msgTx: free=0: head=0,tail=0, used: head=4604d48,tail=4607d3c,wqHead=55941f0,wqTail=55582e0,wqSize=1024
Follow the output for the SNR from each US port
CMTS1#snr0
ifIndex 4
includesContention 0
unerroreds 267012
correctables 1023788
uncorrectables 237651
signalToNoise 155
microReflections 17
equalData 00.A3.FE.86.FF.DA.01.B8.FD.1F.F9.81.3B.E1.00.00.FF.F5.FF.7F.00.E3.FE.A1.FF.61.00.95.FF.29.FF.06
CMTS1#snr1
ifIndex 5
includesContention 0
unerroreds 7957553
correctables 35592
uncorrectables 1922
signalToNoise 248
microReflections 31
equalData FF.F5.FF.F0.00.68.00.38.FE.8B.FF.BB.3F.AF.00.00.00.22.FF.C4.00.2F.FF.D5.FF.E4.FF.E9.FF.FD.FF.E4
CMTS1#snr2
ifIndex 6
includesContention 0
unerroreds 2900338
correctables 123848
uncorrectables 823
signalToNoise 242
microReflections 29
equalData FF.52.00.09.00.CE.00.4D.FE.7B.FF.E2.40.6B.00.00.FF.B6.FF.37.FF.E8.FF.B1.00.82.00.74.FF.D8.00.15
CMTS1#snr3
ifIndex 7
includesContention 0
unerroreds 5593
correctables 63
uncorrectables 720
signalToNoise 0
microReflections 21
equalData FF.FC.FE.0E.01.EE.FE.23.FD.7A.FE.D7.3E.E3.00.00.FE.39.02.02.FF.8B.01.26.00.D2.FF.E2.01.23.01.26
CMTS1#
This is the output for US03 controllers
Cable 0/0 Upstream 3 is up
Frequency 36.000000 MHz, Channel Width 0.800000 MHz, QPSK, Symbol Rate 0.640000 Msps
Spectrum Group: grupo3
SNR 0 dBmv
Nominal Input Power Level 80 dBmV
Ranging Backoff (Start 2, End 8)
Ranging Insection Interval (200 ms)
Tx Backoff Start 2, Tx Backoff End 8
Modulation Profile Group 2
Concatenation is enabled
part_id=0x3137, rev_id=0x3
Range Load Reg Size=0x58
Request Load Reg Size=0xE
Minislot Size in number of Timebase Ticks is = 16
Minislot Size in Symbols = 64
Bandwidth Requests = 0x0
Piggyback Requests = 0x1BE
Invalid BW Requests = 0x0
Minislots Requested = 0x0
Minislots Granted = 0x0
Minislots Size in Bytes = 16, Map Advance (Dynamic): 2495 usecs
UCD Count = 0
Slots 2124, NoUWCollNoEnergy 25, FECCorHCS 39, HCS 39
Req 137107342, ReqColl 41, ReqNoise 42780, ReqNoEnergy 137062787
ReqData 0, ReqDataColl 0, ReqDataNoise 0, ReqDataNoEnergy 0
Rng 1580, RngColl 536870917, RngNoise 843
FECBlks 6376, UnCorFECBlks 720, CorFECBlks 63
CMTS1#
I'm attaching the graphs of historical SNR from the upstream U3 port.
Please I would appreciate any help.
Thanks in advance.
would help to see what the config looks like. asking because you have show controller for upstream 3 only and it's on spectrum group 3. what are the others on? what's the channel width and modulation?
reason I started with fix your return, port 0 has 15.5 SNR and bad micro reflections. port 1 looks good and can support 16Qam, while 0 would have problems with qpsk. so your homework is post a show run (minus the ip's and passwords), output of sh contr cab 0/0 | in SNR (repeat it so you get about 5 outputs to get an average), sh cab modem sum total and get the line tech out to clean up port 0.
are there modems on port 3? connect a modem and do a constant ping to it. then check the snr and controller.
~Carl
Hi cmcaldas, as always thanks for your time.
This is what I have for running config.
hostname CMTS1
!
no service password-encryption
!
!
ip domain-name domain.net.co
ip name-server x.x.x.x.x.
ip name-server x.x.x.x.x
!
no logging control docsis
logging admin-status unconstrained
logging rate-limit 100 1
logging buffered notifications
logging console notifications
logging disable cm_ranging_fail_r103_0
!
!
cable spectrum-group grupo0
band 23000000 40000000
hop action frequency 36000000 priority 2
hop action roll-back
!
cable spectrum-group grupo1
band 23000000 40000000
hop action frequency 36000000 priority 2
hop action roll-back
!
cable spectrum-group grupo2
band 23000000 40000000
hop action frequency 36000000 priority 2
hop action roll-back
!
cable spectrum-group grupo3
band 23000000 40000000
hop action frequency 36000000 priority 2
hop action roll-back
!
interface ethernet 0/0
ip address x.x.x.x x.x.x.x.x
no shutdown
!
interface cable 0/0
ip address x.x.x.x.x 255.255.0.0
ip address x.x.x.x.x. 255.255.0.0 secondary
ip address x.x.x.x.x.x 255.255.255.224 secondary
cable helper-address x.x.x.x.x.x cable-modem
cable helper-address x.x.x.x.x.x host
no shutdown
cable downstream 0 frequency 561000000
cable downstream 0 modulation 64
cable downstream 0 power-level 500
cable downstream 0 rate-limit
no cable downstream 0 shutdown
cable upstream 0 frequency 33000000
cable upstream 0 range-backoff 2 8
cable upstream 0 power-level 80
cable upstream 0 modulation-profile 2
cable upstream 0 rate-limit
cable upstream 0 spectrum-group grupo0
no cable upstream 0 shutdown
cable upstream 1 frequency 33000000
cable upstream 1 range-backoff 2 8
cable upstream 1 power-level 80
cable upstream 1 modulation-profile 2
cable upstream 1 rate-limit
cable upstream 1 spectrum-group grupo1
no cable upstream 1 shutdown
cable upstream 2 frequency 33000000
cable upstream 2 channel-width 1600000
cable upstream 2 range-backoff 2 8
cable upstream 2 power-level 80
cable upstream 2 modulation-profile 2
cable upstream 2 rate-limit
cable upstream 2 spectrum-group grupo2
no cable upstream 2 shutdown
cable upstream 3 frequency 33000000
cable upstream 3 channel-width 800000
cable upstream 3 range-backoff 2 8
cable upstream 3 power-level 80
cable upstream 3 modulation-profile 2
cable upstream 3 rate-limit
cable upstream 3 spectrum-group grupo3
no cable upstream 3 shutdown
cable dynamic-service authorization-mode disable
ip dhcp relay information option
cable flap-list trap-enable
cable downstream schedule priority-wfq
no cable wireless downstream modulation qpsk
!
!
ip route 0.0.0.0 0.0.0.0 x.x.x.x.x
!
!
route-map ruta_nat permit 11
match ip address 101
set ip next-hop x.x.x.x.x
!
snmp-server engineid local 800013750300000000000000
snmp-server context
snmp-server chassis-id Serial Number: XXX-YY-ZZZZ
snmp-server contact
snmp-server sysname CMTS1
snmp-server location
snmp-server trap rate-limit 100 1
!
!
packet-cable
dqos shutdown
no em shutdown
!
!
ipsec
ipsec shutdown
!
!
session-timeout console 0
session-timeout telnet 15
!
telnet session-limit 8
!
alias priv scs show cable modem summary
alias priv snr0 show interfaces cable 0/0 upstream 0 signal-quality
alias priv snr1 show interfaces cable 0/0 upstream 1 signal-quality
alias priv snr2 show interfaces cable 0/0 upstream 2 signal-quality
alias priv snr3 show interfaces cable 0/0 upstream 3 signal-quality
!
!
access-list 101 permit ip 10.4.0.0 0.0.255.255 any
!
no ssh enable
!
logging on
no logging trap
no logging snmp-trap
logging facility local7
!
sntp server 200.91.231.200
clock timezone zone -5 0 daylightsavings on
!
enable rdn-process
!
cable service-class
!
This is the average SNR
SNR average
0- 223,4
1- 248,6
2- 251,4
3- 0
CMTS1#show cable modem summary total
Interface Total Active Registered
Modems Modems Modems
Cable 0/0/U0 26 0 26
Cable 0/0/U1 90 4 86
Cable 0/0/U2 30 1 29
Cable 0/0/U3 2 2 0
Total 148 7 141
CMTS1#
Sorry I didn't understand what you saying about "get the line tech out to clean up port 0" sorry I think it's a matter of translation for me. Please explain me so I can give you the best info possible.
Unfortunaly when I got the sh cab sum total were only 2 modems on port 3. I tried to do ping but no answer.
Thanks in advance.
Question, port 2 and 3 have smaller channel widths than 0 and 1. why is that? have you tried to go to the default of 3.2Mhz wide?
to confirm it's not a port issue, have you moved wires? meaning swap the return between port 2 and 3? if the modems come up on 3 and don't on 2, then it's not the cmts. I would recommend making all the channel widths the same. regarding the tech to clean up port 0, the post you started showed the SNR on port 0 at 15.5 and that's low. for qpsk, it should be 18 and higher.
1. make all the channel width's 3.2Mhz wide
2. swap cables between 2 and 3
~Carl
Hi
I just did what you told me. I changed channel-width port 2 and 3 to 3.2Mhz. The SNR in port 3 still on zero (0). We put those channel-witdh before only for testing purposes but we were aware of that.
Yesterday we did the test swaping the cables, in our case we swap port 1 to 3, and for a moment port 3 were on 230-235 but after seconds dropped to 0, seconds later back higher than 200 and the drop to 0 and so on. We did that test only for a few minutes and we took the decision to reverse the change.
Also we realized yesterday that this behavior was on all ports, the SNR is higher than 200 (20.0) and then dropped to 0. We also did a reload and reload for power reset and port 3 still on 0.
What you telling me about port 0, we are aware that zone has many noise issues so I don't give many importance until the RF guys solve that situation.
I want to believe that port 3 is ok, because it worked at least for a moment, but I don't know how stabilize it so remains higher than 20.
So I resume
1. R/ done without any result
2. R/ we swapped and port goes up for a moment. But later goes "offline"
Hope I explained myself
Thanks in advance.
OK, this is weird. I just saw the historical SNR record for port 3. It seems it worked from 10pm yesterday until 5am from today. I don't know what the heck is going on...
is it possible to combine one or two test modems on port 3? from a pc that can reach the modem network, do a constant ping to that modem, example from ms dos ping 10.10.10.221 -l512k -t
this will have the modem transmit constantly instead of just doing a keep alive or station maintenance. you'll have a carrier there and SNR would be captured. if the modems in the field go out and that test modem stays online, can check with show cable modem time-registered to confirm, would point to a possible node issue
just a thought....
~Carl
Well we have on USport 3, three optical nodes, like 3 branches, so I'm going to disconnect 2 branches and left 1 only connected to USport3. The other branches going to USport2.
Let me do that and I'll post the results.
Thanks for your help
Thanks for posting the results when done. reason why I thought to keep a test modem on is the graph you posted showed there was modems on during the evening. return laser clipping or putting so much noise in the return, could be wiping it out. not sure, it's a thought
Hi cmcaldas. I really appreciate your effort and time. This is what I got.
When I send you the last post I didn't realize that USport3 was up since 20:00hrs yesterday. Now we have multiple cable modems online. So I didn't want to disconnect the branches because we had several calls from our users and playing with their patience. Anyway I just did the ping -t -l and I'm sending you the attachment. This is the outputs:
CMTS1#show cable modem summary total
Interface Total Active Registered
Modems Modems Modems
Cable 0/0/U0 25 0 25
Cable 0/0/U1 90 5 85
Cable 0/0/U2 31 1 30
Cable 0/0/U3 21 1 20
Total 167 7 160
CMTS1#
********
the test cable modem
********
CMTS1#show cable mode time-registered 0017.ee43.5f70
Interface Connect Mac Address Registration Spectrum Group
State Time
Cable 0/0/U3 online 0017.ee43.5f70 000:03:34:36 grupo3
CMTS1#
Average SNR
0- 19.2
1- 23.4
2- 23.0
3- 19.2
Cable 0/0 Upstream 3 is up
Frequency 36.000000 MHz, Channel Width 3.200000 MHz, QPSK, Symbol Rate 2.560000 Msps
Spectrum Group: grupo3
SNR 174 dBmv
Nominal Input Power Level 80 dBmV
Ranging Backoff (Start 2, End 8)
Ranging Insection Interval (200 ms)
Tx Backoff Start 2, Tx Backoff End 8
Modulation Profile Group 2
Concatenation is enabled
part_id=0x3137, rev_id=0x3
Range Load Reg Size=0x58
Request Load Reg Size=0xE
Minislot Size in number of Timebase Ticks is = 4
Minislot Size in Symbols = 64
Bandwidth Requests = 0x0
Piggyback Requests = 0x89979
Invalid BW Requests = 0x0
Minislots Requested = 0x0
Minislots Granted = 0x0
Minislots Size in Bytes = 16, Map Advance (Dynamic): 2495 usecs
UCD Count = 0
Slots 2478114, NoUWCollNoEnergy 17024, FECCorHCS 25537, HCS 25537
Req 1632002333, ReqColl 1502, ReqNoise 285464, ReqNoEnergy 1629777821
ReqData 0, ReqDataColl 0, ReqDataNoise 0, ReqDataNoEnergy 0
Rng 165971, RngColl 536870937, RngNoise 17855
FECBlks 6301630, UnCorFECBlks 266983, CorFECBlks 51719
CMTS1#
I'll attachin the complete output for all US ports in an txt file so the post be short.
How can I make some evaluations to know if is laser clipping?
Thanks in advance.
PD: btw, can you give info about the right SNR levels per modulation and channel-width? is there any table with information for it? website? thanks.
to start, to see if the return laser is clipping, you would need a spectrum analyzer at the head end, you would see additional carriers past the 42Mhz return, not the cable plant guy I use to be. depends on manufacturer of the node as well. have the line tech double check the specs. the show cable modem output shows modems below the configured 8 dbmv receive level. that tells me they are at max transmit level. those will be the ones that have service going on/off during the day and tend to work ok at night. with a 3.2Mhz wide carrier you need a min. of 18 snr for qpsk, 21 for 16 qam and 29 for atdma. shared bandwidth would 5 Mbps for qpsk, 10 for 16 Qam and 15 for 64 Qam. 64 Qam at 6.4 Mhz wide would give you around 32 Mbps. keep in mind that's before docsis overhead.
Hi cmcaldas. All the information provide us has been very important. Unfortunaly our spectrum analyzer is damaged and we are in the purchase process. I'm going to expose RF guys about what you're telling me about the modems below 8dbmv.
Have you ever seen a case like this? I can't make any conclusion until we have our spectrum, but I can "imagine" that the problem is in network, not in the cmts. Could you be agree with me? Your opinion would be very important for me. Hope I'm right because network can be repaired, but cmts is very difficult.
Well, I guess, that's all for the moment. I have to say is working fine, (for now) and when we have our spectrum back, I'll post the the results hoping that is going to help somebody in the group.
Thanks again cmcaldas.
I can say with the information you've posted, the root problem is with the cable plant, not the cmts. this is regardless of the cmts used..
you need a minimum of 18 SNR with qpsk to run without problems. your modems should run with a receive level of 0 +/- 10 ( I know, docsis spec says 15, not 10) and a transmit level of around 45. Use an snmp program to get the information from the modems, so you know what area has the most modem problems and start there. but when a service tech goes to a home to work on a problem, be it a service call to replace a remote control or check a video service, they should also check the modem levels and grounding. bad grounds will cause problems as well. a service call to check internet service via the cable modem needs to have the information of transmit and receive levels on the work order or service ticket so it can be tracked. he shouldn't leave the house if it is out of spec. could be a simple fix as moving the coax to a different port of a 3-way splitter.
the line tech needs to setup the return so a modem that transmit at 45 dbmv hits the cmts receiver with a zero. snr into the cmts should be in the 20's. it may not all be because of the line tech not fixing the plant too. make sure the head end tech hasn't added so much loss in the return that the modems have no choice but to transmit at a higher level, resulting in the node going into distortion or clipping.
you have a 3 part problem and takes 3 groups to resolve.
1. plant needs to be cleaned and aligned
2. head end tech needs to work with line tech so a modem that transmit at 45 hits the cmts with a zero (use a diag modem connected to a node, set with a constant carrier to confirm it's setup correctly)
3. any service call needs to confirm and fix modem level problems and confirm the drop is grounded.
follow these steps and you'll run without issues and be able to move onto docsis 2.0 and 3.0 when the time comes without issues.
~Carl
Hi cmcaldas. I've learned more with you and your posts than any website on internet. I would like to ask you if there's any chance to give us some videoconference and give us some classes to me and RF guys? If is possible please give me a messsage to cabo81 at gmail dot com. If not possible, nevertheless I'll get your steps and put into practice because we've been into so much issues on cable plant.
So I'll be waiting for your message.
Thanks in advance.