[cisco-voip] Special considerations for IAD or PRInetwork-emulation on Cisco hardware...
Nick Matthews
matthnick at gmail.com
Wed May 6 22:47:32 EDT 2009
Hi Jeff,
These are some of the generic echo/audio quality possible causes:
1. Packet loss
-You need to troubleshoot the packet loss in the direction of the
audio problems. In this case if the PBX users are hearing bad voice
quality, that means you need to find out if the packets from the 5800
are being received correctly, not vice versa.
-It's possible your QoS is correct and the packets are dropping in
the provider cloud
-They could be getting dropped in a policer somehwere in your L3/L2 network
-High CPU on any device between gateway and PBX (including the PBX itself)
-Anything more than a few packets can be noticeable.
-Check for interface errors - CRC, input queue, output queue, half
duplex mismatches
2. Jitter
-Much harder to track, because unlike packet loss devices do not
record when they have jitter
-Packet capture is best way to determine
-QoS is major suspect
-Can try ping tests to see if your delay is variable
-High CPU also causes jitter
-Can be in the provider network, albeit less likely than the packet
loss scenario
-Unnecessary devices in call flow like: WAAS, IPS, firewalls
3. Delay
-Adding in DSPs adds ~30-50ms of delay.
-You will start noticing audio problems between 100-150ms, depending
on the jitter.
-GRE/IPSEC tunnels can add software processing delay
4. Hardware
-Bad DSPs
-Physical problems on PBX side or PSTN side - like cables wrapped
around electrical wire
-Bad cards
-Line errors between PSTN : CRC, PCV, LCV, errored seconds
5. Software bugs
-Real bad echo bug around 12.3(11)T7 to early 12.4 mainline (12.4(3c) or so).
-A few other software echo bugs
-A few echo bugs on the IP phones themselves
6. Configuration causes
-Turning off the ecan
-Excessive input gain or output attenuation
-Manually specifying the playout buffer / echo tail
-missing "network-clock-select 1 t1 0/0/0" or similar and slips on line
7. Headsets.
-Enough said.
8. Analog problems
-Signal coming in too hot or soft, requires using input gain or
output attenuation
-Static, clipping, or noise coming from provider line
Happy hunting :)
-nick
On Wed, May 6, 2009 at 5:45 PM, Jeff Anderson <ciscoplumber at gmail.com> wrote:
> I have looked over all of the interfaces for slips and errors and everything
> looks clean. We are the ISP for the customer so I have access to both the
> CPE side and the PE side. In this scenario the the data T1 on the IAD
> recieves clock from the line. The IAD PRI is set to clock source internal.
> The PBX is set to clock source line.
>
> If there is a method to picking IOS, I havent learned it. Our requirements
> for these devices are pretty basic. Any ipvoice feature set should meet our
> requirements.. Does anyone have a preferred IOS version for voice or
> specifically 2431 IAD's.
>
> Thanks,
>
> Jeff
> On Wed, May 6, 2009 at 4:23 PM, Jason Aarons (US)
> <jason.aarons at us.didata.com> wrote:
>>
>> All your questions are around layer1, a show controller or show
>> service-module would indicate if clocking or timing or other layer1 issues.
>> You have to look at both sides. Your side may be clean but the other side
>> dirty.
>>
>>
>>
>> Usually a TTC T-Berd with Y-Cable in the middle would help with analysis.
>> You can even listen in to troubleshoot. You can rent T-Berd for a month for
>> less than $250 from Renfro Test Equipment (Google).
>>
>>
>>
>> Other answers inline
>>
>>
>>
>> From: cisco-voip-bounces at puck.nether.net
>> [mailto:cisco-voip-bounces at puck.nether.net] On Behalf Of Jeff Anderson
>> Sent: Wednesday, May 06, 2009 5:00 PM
>> To: cisco-voip at puck.nether.net
>> Subject: [cisco-voip] Special considerations for IAD or
>> PRInetwork-emulation on Cisco hardware...
>>
>>
>>
>> I have a couple of customers who are experiencing poor audio using an IAD
>> to connect their PBX to our h.323 PSTN GW.
>>
>>
>>
>> Basic Layout.
>>
>>
>>
>> Customer PBX PRI > IAD PRI (Network-emulate) converted to VoIP > MPLS
>> enabled T1 > Core > AS5850 (VoIP converted to TDM) > PSTN.
>>
>>
>>
>> The customer reports that poor audio is experienced inbound into their
>> PBX. They say their outbound audio is okay. This doesnt happen on every call
>> but we do have more than one customer reporting the issue.
>>
>>
>>
>> The audio can be described anywhere from echo, cutting out or distorted
>> audio. The problem has also been reported by users who have 38xx series
>> gateways so i dont think the problem is IAD specific. The AS5850 is a shared
>> device used by many different customers, the majority of which connect using
>> Cisco IP phones and this demographic of customers are not reporting the
>> problem.
>>
>>
>>
>> I have done packet captures on the AS5850 and there is no packet loss
>> reported, the packets arrive in sequence and the jitter is very low.
>>
>>
>>
>> In my testing, I have found double-talk to cause the far end audio
>> (AS5850) to cut-out. All the packets arrive but they contain a payload value
>> of 7F. Some improvement can be made to this condition by disabling
>> non-linear progression.
>>
>>
>>
>>
>>
>>
>>
>> My questions for the community are related to experiences or lessons
>> learned from using PRI network emulate.
>>
>>
>>
>> Can there be too many echo cancellers in the call flow? (i.e. Customer PBX
>> UC500 with Cisco PRI, IAD PRI, AS5850 PRI all with echo cancellers)
>>
>> They are usually local link significant. Bad DSP micro code can cause the
>> issues you are reporting. I’ve done PCM captures that were good yet the DSP
>> turned out to be bad.
>>
>>
>>
>> Does the cable length command have any bearing on voice quality?
>>
>> I haven’t seen the cable length command fix quality issues
>>
>>
>>
>> Are there any special commands i should enter when connecting back to back
>> wth a T1 cable (echo-cancel coverage, nlp, etc)?
>>
>> Who is clocking? Network vs user side? Clock slips, line code violations,
>> etc would create the problems you describe
>>
>>
>>
>> Is a 2 pair sheilded T1 cable necessary for the PRI back to back
>> connection of less than 20 ft?
>>
>> I always use shielded T1, who knows when EMF will be present? A 1 ft cable
>> wrapped around a ballast would be bad..
>>
>>
>>
>>
>>
>> Any insight or suggestions would be greatly appreciated.
>>
>>
>>
>> Thanks,
>>
>>
>>
>> Jeff
>>
>>
>>
>>
>>
>> ________________________________
>>
>> Disclaimer: This e-mail communication and any attachments may contain
>> confidential and privileged information and is for use by the designated
>> addressee(s) named above only. If you are not the intended addressee, you
>> are hereby notified that you have received this communication in error and
>> that any use or reproduction of this email or its contents is strictly
>> prohibited and may be unlawful. If you have received this communication in
>> error, please notify us immediately by replying to this message and deleting
>> it from your computer. Thank you.
>>
>> _______________________________________________
>> cisco-voip mailing list
>> cisco-voip at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-voip
>>
>
>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip
>
>
More information about the cisco-voip
mailing list