Questions

QoS with Nortel IP Phones and Cisco Devices

Tags:
+
0 Votes
Locked

QoS with Nortel IP Phones and Cisco Devices

firenet
Does any one have Cisco switch and/or Cisco Router QoS configurations working with Nortel 2004 IP phones? The configurations that I find are for use with Cisco IP Phones. Nortel IP Phones use a slightly different DiffServ than Cisco Ip Phones and signalling. Thank you,

Firenet
  • +
    0 Votes
    joshgluck

    Can you give a little better example of the environment you are in and where you are looking to deploy QoS. Also, what HW platforms are you running on the cisco side and on the Nortel side (aside from the IP2004s). Is it a CS1KE or IPGW like the 2100? What release, etc.

    I am working on a similar situation here and can share some experiences with our CS1kE + CS1000Bs and IP2004/IP1140s. Due to the design of our network we have not needed to implement QoS to ensure the proper priority to voice when contention occurs up until now. We are deploying a few locations where a CS1kB would be appropriate but do not have the funding. In these cases, the MetroE connection will be carrying a significant amount of data, call signaling and all the RTP streams for 99% of the calls from approx 40 users. This is a significant difference from the configuration we use at our 1000B sites where we provide dedicated PSTN access at each location and the MetroE is only carrying signaling, 5 digit from main, MoH and data. Due to the nature of their calling the majority of network traffic is pushed in and out of the PSTN local to the branch even when the phones are registered to the CS1KE so QoS is not really needed but is a good idea and will be implemented shortly.

    Any more info from your side would be helpful. Also, the Nortel DSCP values are configurable in element manager and pushed to the sets by the sig servers. Do you run the phone system as well as the data network or are you working with the telecom folks jointly.

    ~j

    +
    0 Votes
    firenet

    I am running a Nortel BCM400 and it is located at a Colo. The BCM has 2 PRIs provided by One Communication (formally Conversent). From the colo we have a 10mbs TLS/SES circuit and as a backup connection going to another site where IP Phones are located. The backup connection are two T1s that are bundled and soon to be 4 T1s setup as Multilink as well.
    My Access Layer switch is a Cisco 3750 with PoE.
    The main problem I am having is the IP Phones keep rebooting themselves whenever bandwidth gets to 4 -8 mbs on the TLS circuit(I think this is a what they call Metro Ethernet).This is happening because the signaling is being killed by FIFO.
    Unfortunately I cannot change the DSCP values on the BCM which means that DSCP 40 and 46, signal and media will end up in the same buffer out queue of my switch when I get QoS working. I decided to leave it like this without changing the mappings and just try to get a POlicy to drop excess traffic from these DSCP values trafic is exceeded. Once I have a Policy working then work on changing the DSCP to Cos mappings on the Cisco switch.
    I know I am doing something wrong with my access list from my QoS Policy. I do not see any hit counts increase. I know the BCM listens on Port 7000 and the IP phones on 5000. I probably have to re-arrange my access-list. Do you know the Port range that the IP phones use, it's something like 51000 to 5xxxx? I cannot fine documentation on this for the BCM? Anyway I copied and pasted my Policy access-list any help would be appreciated:
    ip access-list extended VVLAN-ANY
    permit ip 10.154.4.0 0.0.0.255 any log
    permit ip 10.154.4.0 0.0.0.255 any log-input
    ip access-list extended VVLAN-CALL-SIGNALING
    permit udp host 10.154.2.10 eq 7000 10.154.4.0 0.0.0.255 eq 5000 dscp cs5 log
    ip access-list extended VVLAN-VOICE
    permit udp host 10.154.2.10 range 28000 28511 any dscp ef log

    To answer your question "I now inherited the phone system". I don't mind because I'm having fun. But I have already taken live crash course on setting up Target lines, PRI Pools, DN, and DN renumbering, faxing, eCall pilot, ACD group - fun fun.
    Thank you, Chris

  • +
    0 Votes
    joshgluck

    Can you give a little better example of the environment you are in and where you are looking to deploy QoS. Also, what HW platforms are you running on the cisco side and on the Nortel side (aside from the IP2004s). Is it a CS1KE or IPGW like the 2100? What release, etc.

    I am working on a similar situation here and can share some experiences with our CS1kE + CS1000Bs and IP2004/IP1140s. Due to the design of our network we have not needed to implement QoS to ensure the proper priority to voice when contention occurs up until now. We are deploying a few locations where a CS1kB would be appropriate but do not have the funding. In these cases, the MetroE connection will be carrying a significant amount of data, call signaling and all the RTP streams for 99% of the calls from approx 40 users. This is a significant difference from the configuration we use at our 1000B sites where we provide dedicated PSTN access at each location and the MetroE is only carrying signaling, 5 digit from main, MoH and data. Due to the nature of their calling the majority of network traffic is pushed in and out of the PSTN local to the branch even when the phones are registered to the CS1KE so QoS is not really needed but is a good idea and will be implemented shortly.

    Any more info from your side would be helpful. Also, the Nortel DSCP values are configurable in element manager and pushed to the sets by the sig servers. Do you run the phone system as well as the data network or are you working with the telecom folks jointly.

    ~j

    +
    0 Votes
    firenet

    I am running a Nortel BCM400 and it is located at a Colo. The BCM has 2 PRIs provided by One Communication (formally Conversent). From the colo we have a 10mbs TLS/SES circuit and as a backup connection going to another site where IP Phones are located. The backup connection are two T1s that are bundled and soon to be 4 T1s setup as Multilink as well.
    My Access Layer switch is a Cisco 3750 with PoE.
    The main problem I am having is the IP Phones keep rebooting themselves whenever bandwidth gets to 4 -8 mbs on the TLS circuit(I think this is a what they call Metro Ethernet).This is happening because the signaling is being killed by FIFO.
    Unfortunately I cannot change the DSCP values on the BCM which means that DSCP 40 and 46, signal and media will end up in the same buffer out queue of my switch when I get QoS working. I decided to leave it like this without changing the mappings and just try to get a POlicy to drop excess traffic from these DSCP values trafic is exceeded. Once I have a Policy working then work on changing the DSCP to Cos mappings on the Cisco switch.
    I know I am doing something wrong with my access list from my QoS Policy. I do not see any hit counts increase. I know the BCM listens on Port 7000 and the IP phones on 5000. I probably have to re-arrange my access-list. Do you know the Port range that the IP phones use, it's something like 51000 to 5xxxx? I cannot fine documentation on this for the BCM? Anyway I copied and pasted my Policy access-list any help would be appreciated:
    ip access-list extended VVLAN-ANY
    permit ip 10.154.4.0 0.0.0.255 any log
    permit ip 10.154.4.0 0.0.0.255 any log-input
    ip access-list extended VVLAN-CALL-SIGNALING
    permit udp host 10.154.2.10 eq 7000 10.154.4.0 0.0.0.255 eq 5000 dscp cs5 log
    ip access-list extended VVLAN-VOICE
    permit udp host 10.154.2.10 range 28000 28511 any dscp ef log

    To answer your question "I now inherited the phone system". I don't mind because I'm having fun. But I have already taken live crash course on setting up Target lines, PRI Pools, DN, and DN renumbering, faxing, eCall pilot, ACD group - fun fun.
    Thank you, Chris