call recording strangness

Unanswered Question
Aug 13th, 2007
User Badges:

Has anybody around here actually had a successful implementation using Witness to record calls on a Cisco phone system? We have recently implemented it and there seems to be nothing but trouble and the support through Witness is far from stellar. The latest trouble is that when it is recording calls, it will only record about half of the audio. According to the support person at witness all half audio issues are caused by the network. I could see that as a possibility if the folks that are actually on the phone were reporting half audio or missing audio but I have asked several of our agents that have calls which according to Witness are half audio and they have said that there are no issues with the call so I am leaning against it being network related. All that was done to the network to implement witness was setting up a SPAN port. Here is a basic idea of how our network is layed out:

PSTN-->2821 gateways-->3560 Core Switches-->3560 Access Switches-->IP Phones/PC's

I have configured RSPAN for the SPAN port they need to grab the audio:

CORESW01(all of our access switches are connected to ports here)-

monitor session 2 source interface Gi0/1 - 25 rx

monitor session 2 filter vlan 10

monitor session 2 destination remote vlan 500

500 RSPAN

active

CORESW02 (all of our access switches are connected to ports here)-

monitor session 2 source interface Gi0/1 - 25 rx

monitor session 2 filter vlan 10

monitor session 2 destination remote vlan 500

500 RSPAN

active

ACCESSSW01 (this is where the span port is connected)-

monitor session 1 destination interface Gi0/3 - 4

monitor session 1 source remote vlan 500


If anybody has any ideas as to what may be causing our issues I am all ears. Thank you.


  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 3 (1 ratings)
Loading.
paolo bevilacqua Mon, 08/13/2007 - 15:39
User Badges:
  • Super Gold, 25000 points or more
  • Hall of Fame,

    Founding Member

Hi,


I have no experience with Witness software but suppose for a moment that the span port stops forwarding data at some point. The result would be incomplete recordings.

Now I'm not sure how you can verify this hypothesis, perhaps a second span port connected to something different like ethereal o ntop would show you the reliability of the span ports.

Another simpler approach, instead of rspan, could yuo try connecting the recording server directly to the switch where the 2811 is also connected ?

mpozorski Mon, 08/13/2007 - 21:11
User Badges:

Thank you for the reply, there is the possibility that for unknown reasons the span port stops forwarding the data but I find that unlikely because I have a second span port configured and have done captures (using wireshark) and there does not seem to be any lost data but then again I'm not entirely certain as to what the witness software is looking for. I was going to load wireshark or ethereal onto the witness server just to see exactly what it is getting but the vendor will not allow us to upgrade the version of winpcap it is running because it breaks their application. Maybe I can run a capture on the second port for a longer period of time and see what I find from it. The other approach may work but we have 3 2811's that are connected into 2 core switches so RSPAN is the only valid option I have. I originally had it setup that way and they claimed that since the SPAN port was not seeing 'all' of the traffic on the voice vlan it was causing issues with the recorder not seeing the jtapi's from call manager even though call manager was plugged into the same switch (it is network teamed between the core switches). I will give it a shot to attempt to capture the data and see if I can find any thing that would indicate that the SPAN port is acting up. Thanks again for the reply.

Actions

This Discussion