PPP LCP not accepting sent CONFACK

Symptom : User receives the PPP LCP not accepting rcv CONFACK message while accepting an incoming call

Back-7507-R-d#debug ppp nego
PPP protocol negotiation debugging is on
Back-7507-R-d#
*Apr 1 04:53:31: ppp37 PPP: Phase is ESTABLISHING
*Apr 1 04:53:31: ppp37 LCP: I FORCED rcvd CONFACK len 19
*Apr 1 04:53:31: ppp37 LCP: AuthProto PAP (0x0304C023)
*Apr 1 04:53:31: ppp37 LCP: MagicNumber 0x0CFEC631 (0x05060CFEC631)
*Apr 1 04:53:31: ppp37 LCP: MRRU 1524 (0x110405F4)
*Apr 1 04:53:31: ppp37 LCP: EndpointDisc 1 yb (0x1305017962)
*Apr 1 04:53:31: ppp37 LCP: I FORCED sent CONFACK len 21
*Apr 1 04:53:31: ppp37 LCP: MRU 1500 (0x010405DC)
*Apr 1 04:53:31: ppp37 LCP: MagicNumber 0x29B34117 (0x050629B34117)
*Apr 1 04:53:31: ppp37 LCP: MRRU 1500 (0x110405DC)
*Apr 1 04:53:31: ppp37 LCP: EndpointDisc 4 Magic (0x1307043EC61C17)
*Apr 1 04:53:31: ppp37 PPP LCP not accepting sent CONFACK
*Apr 1 04:53:31: ppp37 LCP: O TERMREQ [Closed] id 0 len 4
*Apr 1 04:53:31: ppp37 PPP: Phase is TERMINATING
*Apr 1 04:53:31: ppp37 PPP: Sending Acct Event[Down] id[3C]
*Apr 1 04:53:31: ppp37 LCP: State is Closed
*Apr 1 04:53:31: ppp37 PPP: Phase is DOWN

http://www.ciscotaccc.com/kaidara-advisor/accessdial/showcase?case=K16934604


From the debug traces ,you can see "PPP LCP not accepting sent CONFACK" message on offload server .

configure "ppp mtu adaptive" on virtual-template

modem dialin failure

Symptom
Dialin clients can connect to the access server and modem trainup sounds are heard, but the call is then disconnected

Problem Description
the Network Access Server (NAS) does not launch PPP when it receives a PPP packet from the peer. The connection then times out because the client and NAS cannot negotiate PPP parameters.

Resolution Summary
add "autoselect during-login" and "autoseletc ppp" to line 5/00 5/323 and line 8/00 8/323

voice troubleshooting case debrief-PSTN ring FXS phone failed

Symptom:
Call from PSTN to C3845 FXS analog phone fails, to IP phone is ok.

debug isdn q931 :

Jun 22 01:43:52.430: ISDN Se0/0/1:15 Q931: TX -> CALL_PROC pd = 8 callref =
0x8078
Channel ID i = 0xA98382
Exclusive, Channel 2
Jun 22 01:43:52.434: ISDN Se0/0/1:15 Q931: TX -> PROGRESS pd = 8 callref =
0x8078
Progress Ind i = 0x8188 - In-band info or appropriate now available
Jun 22 01:43:52.462: ISDN Se0/0/1:15 Q931: RX <- DISCONNECT pd = 8 callref
= 0x0078
Cause i = 0x82EF - Protocol error; unspecified

The call from PSTN to FXS will be disconnected by PBX as GW send out a
PROGRESS to it. That means the GW does not accept the call from PBX.

FIX:
configure "voice call send-alert" on your GW to enables the
terminating gateway to send an alert message instead of a progress message
after it receives a call setup message.
-----
more on "voice call send-alert":

On a Cisco 5800 with a PRI-to-PRI configuration or an SS7-to-PRI configuration with the session initiation protocol (SIP) Hairpinning feature and interactive voice response (IVR) enabled, there is no speech path and no ringback for the calls. For calls with IVR and Direct Inward Dialing (DID) disabled, there is speech path
and ringback.
Workaround: Use the global configuration command voice call send-alert. This configures the alerting
message to trigger ringback instead of the progress message.

voice troubleshooting case debrief-dtmf digits

Symptom:
Remote location outbound calls backhaul to main site. When remote site
calls external number and needs enter additional digits after call is connected such as extension number, it appears central site is collecting those digits for another call rather than being forwarded inside call.

Problem:

Unity does not recognize DTMF when calling the PSTN for message notification
When Cisco Unity calls a phone on the Public Switched Telephone Network (PSTN) for message notification, it prompts the called party to enter information through Dual Tone Multifrequency (DTMF). In some cases, Cisco Unity ignores the digits entered and acts as if there was no response. Although this symptom lies with Cisco Unity, the problem is a faulty gateway configuration.
When Cisco Unity makes a call outbound to the PSTN, the DTMF relay must be activated and used for the call.
To activate DTMF on the call, ensure that it matches with an appropriate inbound VoIP dial peer with DTMF relay enabled. If no inbound dial peer is matched, then the default dial peer is used. In this case, DTMF will not be activated for the call, since the default dial peer does not have DTMF enabled.
This example configuration shows this information:
• The VoIP dial peer that services Cisco Unity outbound from the gateway is also used to match inbound by issuing the incoming called-number command.
• The dial string would likely be that used on the appropriate Plain Old Telephone Service (POTS) peer. Alternatively, the incoming called-number . command could be issued to match all dial strings and ensure that DTMF relay is enabled for all calls.

Solution:
• This VoIP dial peer is enabled for DTMF relay by issuing the dtmf-relay h245-alphanumeric command.
dial-peer voice 2 voip
destination-pattern 4...
incoming called-number 9T
!--- This dial string matches the pots dial-peer, or . could be used.
session target ipv4:10.100.25.2
dtmf-relay h245-alphanumeric

voice troubleshooting case debrief-PSTN doesn't send disconnect

topology :

PSTN—gw-IP phone

Symptom:

The IP phone continues to ring after the calling user from PSTN hangs up.

Problem:
PSTN switch doesn’t send disconnect.


Root Cause:

If the local router originates a call out of an FXO voice port, the router has control over that call and can provide the disconnect. If the local router receives a call on an FXO voice port,the router requires that the connected switch provide the disconnect.

Fix:
Use Supervisory Disconnect Tone - (Compatible with Loop Start Signaling only )
If the FXO Supervisory Disconnect Tone feature is configured and a detectable tone from the PSTN or PBX is detected by the digital signal processor (DSP), the analog FXO port goes on-hook.
You can configure the router's FXO voice port to detect supervisory disconnect tones

use the config below for custom cptones and let me know if this works
for you.

voice class dualtone-detect-params 1
freq-max-deviation 25
cadence-variation 25
!
voice class custom-cptone custom
dualtone disconnect
frequency 425
cadence 500 500
!

voice-port 0/1/2
supervisory disconnect dualtone mid-call
supervisory custom-cptone custom
supervisory dualtone-detect-params 1
input gain 14
cptone CN
timeouts call-disconnect 2
timeouts wait-release 2
connection plar opx 13745000
!
------

topology :

PSTN—gw-IP phone

Symptom:

the FXO can not disconnect while the external line hung up.

Problem:
Shangdong NET-COM provide DTMF caller-id signal, which is not China standard "BELLCORE".

the FXO port can not detect the disconnect tone when the external line hangs up.

Root Cause:

If the local router originates a call out of an FXO voice port, the router has control over that call and can provide the disconnect. If the local router receives a call on an FXO voice port,the router requires that the connected switch provide the disconnect.

Fix:
customize tone command
Use Supervisory Disconnect Tone - (Compatible with Loop Start Signaling only )
If the FXO Supervisory Disconnect Tone feature is configured and a detectable tone from the PSTN or PBX is detected by the digital signal processor (DSP), the analog FXO port goes on-hook.
You can configure the router's FXO voice port to detect supervisory disconnect tones

use the config below for custom cptones and let me know if this works
for you.

voice class custom-cptone vc-1
dualtone disconnect
frequency 450
cadence 350 350
!

voice-port 0/1//0
cptone cn
timeouts call-disconnect 1
timeouts wait-release 3
supervisory custom-cptone vc-1
no battery-reversal
!

---

voice troubleshooting case debrief-h245 address

topology :

FAX server----ip2ipGW----TGW---PSTN--FAX1

Symptom:

Fax from FAX1 to FAX server is ok. But from FAX server to FAX1 failed.

Problem:
the Cisco GW doesn't send any H245 address info in neither callproc nor Alerting . XCAPI can not establish media channels.


XCAPI IPIPGW VGW ISDN
----------------------------------------------------------------------------------
H225 setup(H245address) > H225 setup setup
H225calproc(NO H245address) <- H225 callproc callproc
H225alerting(NO H245address) <- H225alerting alerting
??? <- H225connect connect


Root Cause:
This is an integration issue between XCAPI and the IPIPGW.

Normally, as sender (XCAPI) sends h225 setup out, the received (IPIPGW)should return so called H245address (its IP and port for H.245 connection) field in either callproc or Alerting message, and then the sender establishes separate TCP connection to this address as for media channels.

By default, the Cisco GW simply doesn't send any H245address info in neither callproc nor Alerting . XCAPI can not establish media channels.

Workaround:
delete the H245 address when XCAPI send one H225 setup
Fix:
add the h245 address onto IP2IP gw.
Please refer to the commands below:

voice service voip
allow-connections h323 to h323
h323
emptycapability
h225 start-h245 on-connect
h225 h245-address on-connect
call start slow

voice troubleshooting case debrief-connection plar opx

connection plar: on fxo port router goes off hook when ring voltage
detected and places the call to the number configured. If the calling party
hangs up, you can end up with the called phone continually ringing.
- connection plar opx: router places call to configured number when ring
detected but does not go off-hook. Once the called party answers, the fxo
port goes off-hook. If ringing stops (ie calling party hangs up) prior to
the called party answering, call is dropped. Hence no infinite ringing
phone as described above.

When you have the "opx" statement removed, the calling party will hear telco ringback stop immediately when our FXO port seizes the line, but called party hasn't answered yet. In effect, two stage dialing. This may annoy some callers when the called party never answers but billing has begun because the telco thinks the called party has answered. "OPX" fixes this problem, but of course you have some other problem to debug in your situation ...

voice troubleshooting case debrief-call drop after 5 mins

Topology
IP phone > Cisco CallManager > H.323 Voice Gateway > Public Switched Telephone Network (PSTN)
Symptom:
PSTN calling IP phone gets dropped after 5 minutes.

Problem description:
CCM and H323 has keepalives sent between them, 1 every minute, after the 4 keepalive is missed, call will be dropped.

The messages below are the output of “debug isdn q931” and “debug ip tcp transaction”:

Mar 16 19:45:17.966: TCP0: keepalive timeout (1/4)
Mar 16 19:46:17.966: TCP0: keepalive timeout (2/4)
Mar 16 19:47:17.966: TCP0: keepalive timeout (3/4)
Mar 16 19:48:17.966: TCP0: keepalive timeout (4/4)
Mar 16 19:48:17.966: TCP0: state was ESTAB -> CLOSED [40523 ->
192.168.17.244(1720)]---CCM IP address
Mar 16 19:48:17.966: TCB 0x46666038 destroyed
Mar 16 19:49:13.478: ISDN Se0/2/1:15 Q931: RX <- DISCONNECT pd = 8
callref = 0x0201
Cause i = 0x8090 - Normal call clearing
Mar 16 19:49:13.478: %ISDN-6-DISCONNECT: Interface Serial0/2/1:9
disconnected from 01085155315 , call lasted 415 seconds
Mar 16 19:49:13.482: ISDN Se0/2/1:15 Q931: TX -> RELEASE pd = 8 callref
= 0x8201

They indicate that after 4 keepalive packets are missed between the
Callmanager server and the voice gateway,the call gets dropped.

Fix:
disable the Trend Micro firewall service on the Cisco CallManager server

voice troubleshooting case debrief-no ringback tone

Problem description:
No ringback tone when calling from PSTN

Explanation:
- if the AS5300 sends a "180 Ringing", the IP phone is expected to play local ringback.
- if the AS5300 sends a "183 Session Progress", then it will cut-through early-media and the ringback will come from the PSTN switch.

Now there can be some scenarios, depending on the PI value that is send in the ISDN messaging, where we have to tweak the AS5300 behavior.

Action Plan:
1.Please capture the following debug while making test calls:
"debug ccsip all" and "debug isdn q931".
2.you may be able to add "no isdn incoming alerting add-PI" under the serial interface in order to get the 180 ringing.Please try it and let me know how it goes.
3.You can play around with the PI settings / mappings on the 5300:
http://www.cisco.com/univercd/cc/td/doc/product/software/ios121/121newft/121t/121t5/dt4t323.htm
1- Please check the doc below.It explains well how to fix this problems when ringback tone is not heard:
http://www.ciscotaccc.com/kaidara-advisor/voice/showcase?case=K78361372

Try to configure the two commands mentioned in the article above

progress_ind alert enable 8
progress_ind setup enable 3

under the POTS dial peer towards the PSTN .

2- We can generate the ring back tone locally in the GW by typing the command
'tone ringback alert-no-pi'
under the SIP dial peer that the call comes in on the GW.

Please add this hidden command:
progress_ind alert strip 2

" progress_ind setup enable 3":
The outbound call should include a PI of 3, indicating that the originating device is non-ISDN, and that ringback should be provided inband by the PSTN. The progress_ind setup enable 3 command ensures this behavior.
"progress_ind alert enable 8":
The returning alert message from the PSTN should then include a PI of 8, indicating to the originating gateway that inband information (in this case ringback tone) follows, and that the call should be cut through in the reverse direction so that this information can be heard at the origin.
If the returning alert message does not include a PI of 8, the progress_ind alert enable 8 command can be inserted into the Plain Old Telephone Service (POTS) dial peer and will make the originating Cisco IOS gateway cut through the voice path, as if it received an alert with a PI of 8 from the PSTN.
"tone ringback alert-no-pi "
The gateway is also capable of providing ringback tone on the IP call leg if necessary. The tone ringback alert-no-pi command will enable this functionality, and should be used only if the PSTN switch cannot provide inband ringback tone.

voice troubleshooting case debrief-FXO ports hang

Problem description:
FXO ports hang after PSTN to PSTN (hairpin) call forwarding disconnected by the calling party

Symptom:
When a call comes from the PSTN, the call is been answered and the user transfer it to another user through the PSTN line ( either to mobile phone or landline) . The established call occupies two lines (ports) on the CCM router. After the external parties hang up the lines, the router doesn't release the particular lines for further use.

Explanation:
On a transferred call the PSTN side is the originating side.If this side hangs up first without signaling the FXO on the CCM that the call is disconnected, CCM has no way to realize that the FXO ports should be released.

Action Plan:
1.applying the following commands to the voice-ports
supervisory disconnect dualtone mid-call
timeouts call-disconnect 5
timeouts wait-release 5
(this step has been done but it doesn't fix).
2.Please get the cadence and frequency of the disconnect tone from the telco.
3.Create a custom cptone if you can get the frequency and cadence from the telco and apply it to the voice port.
Sample configuration :
e.g. if the telco say the frequency and cadence they use for the tone disconnect are n1,
and n2 n3,respectively, then this what you will need to add:
voice class custom-cptone c-jason
dualtone disconnect
frequency n1
cadence n2 n3

Workaround:
If you can't get the required disconnect tone parameters from the telco,the only workaround is either reconnecting the cable from the FXO port to the or shut/no shut the voice-ports to release the engaged ports.

voice troubleshooting case debrief--CUE

Problem description:
CUE, by default, directs a caller to the operator's extension as defined by "voicemail operator's extension" parameter after they leave a message, press pound (#),and mark it as either normal priority or urgent.

How do I dictate where it sends the callers in that 1 situation- not on a system-wide level?

I realize I can issue a "no voicemail conversation caller recording-prompt" and remove that option from at least being presented to the caller, but that's a band-aid, not a fix.

I can suppress the message instructing the user how to mark their message as urgent, but it doesn't really change the fact that it's attempting to transfer an outside caller to the operator after having left a message and marking it manually.



Solution:
Please configure the voicemail zero-out number with a number that does not exist and each individual's mailbox zero-out number can be configured to the real operator.
This allows the caller to press zero and reach the opertor, but if nothing is entered,they will be transfered to a non-existant extension, which will cause a disconnect to be sent.

voice troubleshooting case debrief--SIP

Topology
SIP phone-AS5300-E1-PSTN

Symptom:
No ringback tone when calling from PSTN

Explanation:
- if the AS5300 sends a "180 Ringing", the IP phone is expected to play local ringback.
- if the AS5300 sends a "183 Session Progress", then it will cut-through early-media and the ringback will come from the PSTN switch.

Now there can be some scenarios, depending on the PI value that is send in the ISDN messaging, where we have to tweak the AS5300 behavior.

Fix:
generate the ring back tone locally in the GW by typing the command
'tone ringback alert-no-pi'
under the SIP dial peer that the call comes in on the GW.

"tone ringback alert-no-pi "
The gateway capable of providing ringback tone on the IP call leg if necessary. The tone ringback alert-no-pi command will enable this functionality, and should be used only if the PSTN switch cannot provide inband ringback tone.