Analog Aastra phones on VG224 disconnect calls when going off-hook

[mmertens]

I have a large IPT deployment (CUCM 6.1) with 600 IP phones and 400 analog devices. The analog devices mostly consist of Aastra analog phones hung off of 25 VG224s. All phones are configured as SCCP, including the Aastra analog end-points on the VG224 (running 12.4(20). Intermittently, when an Aastra analog phone user goes off-hook to answer a call, the call will disconnect. If the VG224 analog end-point is changed to MGCP (and the other analog end-point making the call is changed to MGCP), then the problem does NOT occur. I have not been able to verifiy whether it occurrs when an IP phone calls an analog Aastra phone.

  Also, if the Aastra phone is swapped to another manufacturer (such as Cortelco), the problem also goes away. The problem is: we need features in SCCP, so changing 300 analog end-points to MGCP is not a solution; and swapping out 300 analog phones from Aastra - Cortleco (or whoever) is not a solution. There must be some kind of feature/timer, incompatibility???!!!

  I've been playing with timers, upgraded VG224 code, swapped tip/ring......the problem sometimes happens after 1-5 calls, or 50-100 calls, but certainly averages around 1/20. The customer is quite, quite annoyed at this point.

THANKS!!!!!

[Aaron Harrison]

Hi

What debugging/tracing have you tried so far?

I would probably set up tracing of SCCP at the detailed level in CCM, and possibly also something like ' debug voip application stcapp' and log it out to syslog.

Then get your users to log occurrences (exact time, called and calling number), and when you get one trace that call back. You should be able to see whether or not the phone is triggering some kind of supplementary service or other as this will be signaled out to CCM.

Regards

Aaron

[mmertens]

Aaron, thanks for responding.....we've done traces and Cisco TAC had verified that the

VG224 port is receiving a signal on the analog side that it interperits as the Aastra going back on-hook. Ideally, I'd love to know the difference in MGCP vs. SCCP end-point disconnect timeouts, as that is where my suspicion is.......I would guess that the Aastra phone produces some transient voltage or something durinng the off-hock process (or it's border-line to what SCCP interperets as going off-hook). Aastra states they have a million of these phones across the planet and have never heard of the problem, and the fact that SCCP/MGCP (Cisco) or changing phone models (from Aastra) just points to some slight incompatibility issue.

I would think someone else in the world has run into this....

Thanks.

[Aaron Harrison]

Hi

Well - there are more 'things' you can do with SCCP.

As I understand it, a quick on hook is a 'hookflash', and that triggers a transfer by default. If it's a slightly longer on-hook, then it's hanging up.

How long is the onhook period?

Aaron

[mmertens]

Aaron,

   Below are all my voice port parameters: I really though the Supervisory Disconnect timer would have been key so I increased that by using the batter-reversal command with no effect. I've also modified both the hookflash-in max and min values without help. I think my next step is just to look a the timers of an MGCP-controlled analog end-point and hopefully see some differences.....I hate just blindly (somewhat) changing parameters, like I've been doing....Thanks for your continued input.

Mike.

Foreign Exchange Station 2/0 Slot is 2, Port is 0
Type of VoicePort is FXS
Operation State is DORMANT
Administrative State is UP
The Last Interface Down Failure Cause is Administrative Shutdown
Description is not set
Noise Regeneration is enabled
Non Linear Processing is enabled
Non Linear Mute is disabled
Non Linear Threshold is -21 dB
Music On Hold Threshold is Set to -38 dBm
In Gain is Set to 0 dB
Out Attenuation is Set to 0 dB
Echo Cancellation is enabled
Echo Cancellation NLP mute is disabled
Echo Cancellation NLP threshold is -21 dB
Echo Cancel Coverage is set to 64 ms
Echo Cancel worst case ERL is set to 6 dB
Playout-delay Mode is set to adaptive
Playout-delay Nominal is set to 60 ms
Playout-delay Maximum is set to 1000 ms
Playout-delay Minimum mode is set to default, value 40 ms
Playout-delay Fax is set to 300 ms
Connection Mode is normal
Connection Number is not set
Initial Time Out is set to 15 s
Interdigit Time Out is set to 10 s
Call Disconnect Time Out is set to 60 s
Supervisory Disconnect Time Out is set to 750 ms
Ringing Time Out is set to infinity
Wait Release Time Out is set to 30 s
Companding Type is u-law
Region Tone is set for US

Analog Info Follows:
Currently processing none
Maintenance Mode Set to None (not in mtc mode)
Number of signaling protocol errors are 0
Impedance is set to 600r Ohm
Analog interface A-D gain offset = -3.0 dB
Analog interface D-A gain offset = -3.0 dB
FXS idle voltage set to high
alt-battery-feed set to feed1
Ring DC offset set to 0 volt
Station name None, Station number None

Caller ID Info Follows:
Standard BELLCORE
Output attenuation is set to 14 dB
Caller ID is transmitted after 1 ring(s)
Translation profile (Incoming):
Translation profile (Outgoing):

Voice card specific Info Follows:
Signal Type is loopStart
Ring Frequency is 25 Hz
Hook Status is On Hook
Ring Active Status is inactive
Ring Ground Status is inactive
Tip Ground Status is active
Digit Duration Timing is set to 100 ms
InterDigit Duration Timing is set to 100 ms
Hookflash-in Timing is set to max=1000 ms, min=150 ms
Hookflash-out Timing is set to 400 ms
No disconnect acknowledge
Ring Cadence is defined by CPTone Selection
Ring Cadence are [20 40] * 100 msec
Ringer Equivalence Number is set to 1
VGWAT012#config

[Aaron Harrison]

Hi Mike

OK - so for hookflash, you have min and max timers.

Less than min = do nothing

>min, but max = hang up

So do you know the length of the on hook period?

If you set the min to longer than the on hook period, and max to whatever is the longest you can set it to, then you may get results.

Incidentally, what commands are you using to change the hookflash? I thought you could only change the max timer?

RE: battery reversal - this detects a hangup, usually more applicable to FXOs. If this is what's being seen, then maybe 'no battery-reversal' will stop that function.

Aaron