Problem with loss of frame every two hours

cisco.noob · ‎02-03-2008

Hi, I have been having an annoying problem with my internet connection since I changed provider. The problem is that every 2 hours (almost exactly) the line goes down. I turned on some atm logging, and this is what happens:

000060: *Jul 12 13:21:57.947 UTC: DSL: Defect: LOF: retraining

000061: *Jul 12 13:22:00.447 UTC: Changing vc 0/34vc-state to ATM_VC_SHUTTING_DOWN

000062: *Jul 12 13:22:00.447 UTC: Changing vc 0/34vc-state to ATM_VC_NOT_IN_SERVICE

000063: *Jul 12 13:22:02.447 UTC: %LINK-3-UPDOWN: Interface ATM0, changed state to down

000064: *Jul 12 13:22:02.447 UTC: PVCD(ATM0): clearing event queue

000065: *Jul 12 13:22:02.447 UTC: PVCD: ATM0 Forgetting discovered PVCs...

000066: *Jul 12 13:22:02.447 UTC: PVCD: Removing all dynamic PVCs on ATM0

000067: *Jul 12 13:22:02.447 UTC: PVCD: Restoring MIXED PVCs w/ default parms on ATM0

000068: *Jul 12 13:22:02.447 UTC: PVCD: Marking static PVCs as UNKNWN on ATM0

000069: *Jul 12 13:22:02.451 UTC: PVCD: Marking static PVC 0/34 as UNKNWN on ATM0 ...

000070: *Jul 12 13:22:03.447 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface ATM0, changed state to down

000071: *Jul 12 13:22:04.447 UTC: %LINK-3-UPDOWN: Interface BVI1, changed state to down

000072: *Jul 12 13:22:05.447 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface BVI1, changed state to down

000073: *Jul 12 13:22:24.143 UTC: %LINK-3-UPDOWN: Interface ATM0, changed state to up

000074: *Jul 12 13:22:24.683 UTC: ATM(ATM0.1):point-to-point interface does not have a VCD

000075: *Jul 12 13:22:25.143 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface ATM0, changed state to up

000076: *Jul 12 13:22:26.143 UTC: %LINK-3-UPDOWN: Interface BVI1, changed state to up

000077: *Jul 12 13:22:26.203 UTC: Changing vc 0/34 vc-state to ATM_VC_ESTABLISHING_VC

000078: *Jul 12 13:22:26.203 UTC: Changing vc 0/34 vc-state to ATM_VC_NOT_VERIFIED

000079: *Jul 12 13:22:26.203 UTC: Changing vc 0/34 vc-state to ATM_VC_UP

000080: *Jul 12 13:22:27.143 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface BVI1, changed state to up

It seems that there is a problem with my connection. Indeed, to be fair, my connection is not arranged in the best way. I couldn't find the place where the telephone line comes in (I think it's hidden beyond the ceiling), so I couldn't install the splitter in the right way. So I decided to use it as a filter and connect it just to my phone.

That said, what I don't understand is

1) why the problem started after I changed provider

2) why I'm not having this problem with my old router

3) why, although I don't have the problem every exactly 2 hours (it's more around 2 hours and 3-6 minutes), it's still happening quite regularly

Also, if the problem is the line, considering that I have problems in finding the right spot for the splitter, would it be okay to use 2 splitters, one for the phone and one for the router?

configuration is attached. Any help appreciated. Thanks

Ops, I forgot to mention the model of my router: it's a 877W

paolo bevilacqua · ‎02-03-2008

Hi,

Pls upgrade your dsl firmware, from:

ftp://ftp.cisco.com/pub/access/800

the one you want is 3.0.33, and there is a doc file expalining how to install it.

Hope this helps, please rate post if it does!

cisco.noob · ‎02-03-2008

Hi, thanks for the reply.

I already upgraded it. I did some research before posting :-)

show dsl interface atm 0 says:

Alcatel 20196 chipset information

ATU-R (DS) ATU-C (US)

Modem Status: Showtime (DMTDSL_SHOWTIME)

DSL Mode: ITU G.992.1 (G.DMT) Annex A

ITU STD NUM: 0x01 0x1

Vendor ID: 'ALCB' 'BDCM'

Vendor Specific: 0x0000 0x6193

Vendor Country: 0x00 0xB5

Capacity Used: 72% 92%

Noise Margin: 15.0 dB 7.0 dB

Output Power: 20.0 dBm 11.5 dBm

Attenuation: 27.5 dB 10.5 dB

Defect Status: None None

Last Fail Code: None

Watchdog Counter: 0x75

Watchdog Resets: 0

Selftest Result: 0x00

Subfunction: 0x00

Interrupts: 8223 (0 spurious)

PHY Access Err: 0

Activations: 4

LED Status: ON

LED On Time: 100

LED Off Time: 100

Init FW: init_3.0.33_nobist.bin

Operation FW: AMR-3.0.033.bin

FW Version: 3.033

in the meanwhile, my router went down at:

13:21

15:33

17:36

(the clock is not set, so time doesn't match; it's just to show how regularly it happens)

what I noticed is that my provider has a lease time for dhcp of 2 hours, but there appear to be no correlation.

scottmac · ‎02-03-2008

I think the issue is likely the 7.0db Noise Margin.

Try another / some other cables to your demarc.

It could be that your old modem / router has better specs relating to selectivity and sensitivity, that's why it worked OK.

Two splitters will make the problem worse (it will attenuate / reduce the signal even more).

Also check that your system is properly ventilated and permits good airflow all the way around ... it could be overheating and shutting down (how long does it stay down for?)

It could be a weak power cube (heats up, shuts down ... make sure it also has good air flow)

I'm gonna stick with the cabling though, that's the most likely culprit with the information given.

If that doesn't improve things, call your provider and ask them to test the line. Explain that it's an intermittent issue and may need some prolonged diagnostics.

See if they can / will "pull PM Stats" for 24 hours. PM = "Performance Monitoring" ... it's something they absolutley CAN do (it's part of the basic monitoring suite they would use to manage their DSL network), it'

s just a matter of whether they WILL do it.

If the Tier 1 says no, get an escalation to Tier 2 or 3 and push it.

If the PMs come up normal, ask them to extend your DHCP lease to three hours ... I had a customer with a complete interface down/up/negotiate cycle at every lease renewal (every HOUR). If the time period extends to three hours, you have your cause.

Here are my stats for a 3 Meg DSL from AT&T to a 2811 and DSL / POTS WIC:

ISR#show dsl interface atm0/1/0

Alcatel 20150 chipset information

ATU-R (DS) ATU-C (US)

Modem Status: Showtime (DMTDSL_SHOWTIME)

DSL Mode: ITU G.992.1 (G.DMT) Annex A

ITU STD NUM: 0x01 0x1

Vendor ID: 'ALCB' ' '

Vendor Specific: 0x0000 0x0000

Vendor Country: 0x00 0x00

Capacity Used: 58% 60%

Noise Margin: 14.5 dB 19.0 dB

Output Power: 16.0 dBm 12.0 dBm

Attenuation: 46.5 dB 26.5 dB

Defect Status: None None

Last Fail Code: None

Watchdog Counter: 0x8E

Watchdog Resets: 0

Selftest Result: 0x00

Subfunction: 0x15

Interrupts: 5309 (0 spurious)

PHY Access Err: 0

Activations: 4

LED Status: ON

LED On Time: 100

LED Off Time: 100

Init FW: embedded

Operation FW: embedded

FW Version: 3.8131

Good Luck.

Scott

cisco.noob · ‎02-04-2008

If it's the network, and a splitter or a filter won't help, I'm afraid I'm in trouble, because I can't ask anyone to come and check the line if I can't find the spot where the line comes in, which might be the root of the problem.

What would be a good value for the margin?

Thanks for the info.