877W intermitant DSL loss

[azaghul33]

I am running an 877W in the home that for the most part works fine, however after being connected for a variable amount of time I partially loose my DSL connection. Line sync stays up, but the traffic is lost.

A shut/no sut of the atm or dialer interfaces does not resolve the issue, only a reload or power down/up will fix the prob short term till it happens again.

Am currently running advipservicesk9-mz.151-3.T1, but have also rolled back to advipservicesk9-mz.124-24.T5 which has the same issue. Have tried a 857W with the same results. So ruling out IOS and hardware, its pointing to a config issue. It has also only been an issue since changing from ADSL1 to ADSL2+.

interface ATM0

no ip address

no shutdown

no atm ilmi-keepalive

exit

interface ATM0.1 point-to-point

pvc 8/35

  pppoe-client dial-pool-number 1

  exit

exit

!

interface Dialer0

ip mtu 1452

ip address negotiated

ip nat outside

dialer pool 1

dialer-group 1

encapsulation ppp

dialer idle-timeout 120

dialer fast-idle 20

ppp authentication chap callin

ppp chap hostname <hostname>

ppp chap password <password>

exit

!

interface BVI1

ip nat inside

exit

!

bridge irb

bridge 1 protocol ieee

bridge 1 route ip

!

interface BVI1

ip address 192.168.1.1 255.255.255.0

ip virtual-reassembly

exit

!

interface Vlan1

bridge-group 1

bridge-group 1 spanning-disabled

exit

000066    "ATM: PVC removed, ATM0.1 VCD 1 (8/35)"

000067    "Vi1 PPP: Missed 5 keepalives, taking LCP down"

000068    "Vi1 PPP DISC: Missed too many keepalives"

000069    "PPP: NET STOP send to AAA."

000070    "Vi1 IPCP: Event[DOWN] State[Open to Starting]"

000071    "Vi1 IPCP: Event[CLOSE] State[Starting to Initial]"

000072    "Vi1 CDPCP: Event[DOWN] State[Stopped to Starting]"

000073    "Vi1 CDPCP: Event[CLOSE] State[Starting to Initial]"

000074    "Vi1 LCP: O TERMREQ [Open] id 2 len 4"

000075    "Vi1 LCP: Event[CLOSE] State[Open to Closing]"

000076    "Vi1 PPP: Phase is TERMINATING"

000077    "Vi1 LCP: Event[DOWN] State[Closing to Initial]"

000078    "Vi1 PPP: Phase is DOWN"

000079    "NTP Core (INFO): system event 'event_clock_reset' (0x05) status 'sync_alarm, sync_unspec, 8 events, event_peer/strat_chg' (0xC084)"

000080    "NTP Core (NOTICE): Clock synchronization lost."

000081    "Di0 Deleted neighbor route from AVL tree: topoid 0, address 220.233.1.45"

000082    "Di0 IPCP: Remove route to 220.233.1.45"

000083    "%DIALER-6-UNBIND: Interface Vi1 unbound from profile Di0"

000084    "%LINEPROTO-5-UPDOWN: Line protocol on Interface Virtual-Access1, changed state to down"

000085    "%LINK-3-UPDOWN: Interface Virtual-Access1, changed state to down"

I've tries some changes mentioned here and from other sites, but no success yet...

Any sugestions appreciated.

[billobob123]

Hi

Please do a show tech when all is good and a sh tech when it is not this will remove some user names and passwords in the process but check it as well and post the results.

Do you get the same problem with the ISPs modem ie not your cisco.

B

Sent from Cisco Technical Support iPad App

[azaghul33]

Hi,

I've attached the normal running and down sh tech logs as requested. Though my ISP does not supply preconfigured modems, we can buy a modem from them and configure it with supplied/online instrustions, or supply our own. My D-Link had been up for over a week (purchased to prove an ISP fault recently), while the 877 went down about 4 times yesterday. This problem has been on going for a few months, but as I've only been home the occasional weekend in the past 6 months I could (sort of) live with it.

Cheers, Dave

[billobob123]

thanks,

im looking at the files now, youve got an interesting issue, but perhaps we can get to the bottom of what is going on.

let you know

[billobob123]

hang on, whats the time on your router? make sure that it is right and if not a network clock would help.

you have this

000078    "Vi1 PPP: Phase is DOWN"

000079     "NTP Core (INFO): system event 'event_clock_reset' (0x05) status  'sync_alarm, sync_unspec, 8 events, event_peer/strat_chg' (0xC084)"

see if setting the clock sorts out the problem,

[azaghul33]

the router is configured for local time, sync'd via ntp, sh clock is giving the correct local time.

Janus#sh ntp ass

  address         ref clock       st   when   poll reach  delay  offset   disp

*~192.189.54.17   130.95.179.80    2     30     64   377 39.033  -6.827  4.143

+~121.0.0.42      209.81.9.7       2     39     64   377 40.461   0.840  4.099

+~150.101.72.73   203.12.160.2     3     53     64   377 47.374   2.156  3.948

* sys.peer, # selected, + candidate, - outlyer, x falseticker, ~ configured

Cheers, Dave

[azaghul33]

thanks, any assistance appreciated

[Leo Laohoo]

Can you please post the complete output to the command "sh dsl int atm" please?

[azaghul33]

sh dsl int atm 0 attached

[Leo Laohoo]

Your downstream (DS) Noise Margin is too low (value 3.5 dB).  The higher the value the better.

The upstream (US) is so-so.  How FAR are you from the exchange?

Cisco Show DSL info explained

[azaghul33]

as the crow flies, about 900m to the exchange, allowing for a less direct route, probably about 1250m

[Leo Laohoo]

Hmmmm ... you need to get your noise margin checked.  DO you have an ADSL splitter installed on the phones/fax machines/intruder alarm?

[azaghul33]

I have an ADSL filter installed on the fall-back phone line when not using SIP

[billobob123]

had a similar problem on my line see with your ISP about turning down the speed a little doesn't need to be much it could sort out your dropping issues it did for me, okay I gave upa Meg but the line instead of cutting out every twenty minutes is rock solid

Sent from Cisco Technical Support iPad App

[azaghul33]

i had the same prob with speed on my old ADSL1 link, it got flakey and would fully reset the link (carrier sync lost) anywhere between 7.2 and 7.5Mb/s, kept my d/l speed to 7.0Mb/s or under and I was fine

however this prob will keep carrier sync, but drop connection just doing normal web browsing/mail retreval, no heavy d/l usage. had a connection loss earlier today, just normal pc background activity, no usage at all....weird...

any debug logs I should be looking at more closely to determing why the ATM int is going down?