DHCP Parse Error

It would seem a great many people are running into this error so I thought I would post a quick article that would help all those looking for a solution to their problem. The Nortel i2002/i2004 (any Nortel IP phone) will return the error “DHCP parse error” if the DHCP server returns DHCP option 128 with an invalid format.

In one of my first blog posts entitled “DHCP Options (VoIP)” I outlined the DHCP options and the format that the phone expected. Kenneth quickly pointed out that I had fat-fingered the example DHCP string omitting the semicolon between the values for the primary Call Server (S1) and the secondary Call Server (S2). (I’ve since fixed the original post)

Here’s the format the DHCP option code 128 which the IP phones expects to be returned from the DHCP server when the IP phone is configured for “Full DHCP”;

Nortel-i2004-A,iii.iii.iii.iii:ppppp,aaa,rrr;iii.iii.iii.iii:ppppp,aaa,rrr.

Where the following values are;

  • “Nortel-i2004-A” = Option #128 begins with this string for all Nortel IP phones
  • “iii.iii.iii.iii” = the IP Address of the Call Server (S1 or S2)
  • “ppppp” = port number for the Call Server
  • “aaa” = the Action for the Server
  • “rrr” = the Retry Count for the Server

The IP Address must be separated from the port number by a colon (:). The parameters for the Primary (S1) and the Secondary (S2) Call Servers are separated by a semicolon (;). The string must end a period (.).

If you are using an ISC DHCP server (this is generally included with any of the Linux distributions) you can refer to my article entitled “ISC DHCP Server (Nortel VoIP)” for information on how to setup and configure the ISC DHCP server to work with Nortel IP Telephony.

Cheers!

Update: Wednesday January 28, 2009

I just came across a “DHCP Parse Error” that actually had nothing to-do with the DHCP string. We had a phone that was using the same DHCP template as 9 other i2002 phones but it was displaying “DHCP Parse Error” when it booted up (this only surfaced after we upgraded to 0604DCG). The phone was actually configured incorrectly. It was suppose to be configured for a Voice VLAN and configured to get the Voice VLAN via LLDP-MED. The LLDP-MED option had been left set at “0″ (disabled) so the phone was trying to make a DHCP request looking for Option 191 while the DHCP server was only configured to return Option 128 as detailed above. It only took a few minutes to fix but I can see how someone could spend hours stuck on this.

Cheers!

{ 23 comments… add one }

  • Ben November 24, 2008, 3:23 am

    Hi Michael,
    Thanks for your valuable info !! :)

    Currently i am having this problem where if connecting to Switch 2950, it can’t work, but to Switch 3560, it can work. Have you face with this problem before?

    Thanks !

  • Sajid Qureshi April 8, 2009, 11:41 am

    Hi,

    We just recently bought Nortel 1110 series IP Phones to connect with our IP PBX system. But when I connected the IP phone to LAN, it was displaying DHCP parse error. Kindly anybody tell me the simple solution to fix this problem.

    Thanks in advance!

    Sajid

  • Michael McNamara April 8, 2009, 6:44 pm

    Hi Sajid,

    I believe the article above points out several possible problems that can lead to a “DHCP Parse Error”.

    It’s assumed that you’ve configured the phone, switch and DHCP server appropriately.

    Good Luck!

  • Sajid Qureshi April 16, 2009, 1:14 pm

    Hi Michael,

    Thanks for cmment!

    As I mentioned in my previous post that we have just bought new Nortel IP Phones to use with our existing IPPBX. Before we were using diferent IP Phones and those are working properly. Nortel phone recently we bought and when I tried to configure them and connected them to our LAN immidiately they are displaying message “DHCP Parse Error”

    So advice me, can I configure the Nortel IP Phones with any 3rd party IP-PBX? and How I get rid of this “DHCP Parse Error”. I read your posts but I couldn’t get them. :( Just please advice me how can I confgure Nortel Ip Phones with any 3rd party system?

    I shall be thankful!

    Regards,

    – SAjid

  • Michael McNamara April 16, 2009, 10:52 pm

    Hi Sajid,

    Nortel IP phones will only work with Nortel based Call Servers, Succession 1000 (CS1K) or Business Communications Manager (BCM), out of the box. With that said, depending on the model of your phone you can load SIP firmware onto an 1120e/1140e phones can get them to connect to any SIP compliant solution.

    Good Luck!

  • Sajid Qureshi April 17, 2009, 7:09 am

    Hello Michael,

    Thanks for continuous guideline!

    Our Nortel IP-Phones model is:NTYS02, Series: 1110. As you told, by loading SIP firmware we can use these IP Phones with any SIP softswitch. right? If so, what is solution to get the interface of IP-Phone to configure it with any SIP server?

    Regards,

    – Sajid

  • Michael McNamara April 18, 2009, 11:08 pm

    Hi Sajid,

    I’m not 100% sure that there is a SIP firmware available for that model.

    Have a look at the comments in this post and you’ll find the various links that contain the information your looking for.

    Good Luck!

  • Gord April 19, 2009, 8:30 am

    There is currently no SIP firmware load for the IP Phone 1110/1150E. It is available for the IP Phone 1120E/1140E and includes support for the Expansion Module.

    In terms of the “DHCP Parse Error”, advice is to:
    - check the configuration of the DHCP server
    - ensure that the latest firmware is loaded on the IP Phone. Phones shipped from the factory may NOT have the most recent version so the first step should always be to upgrade them.

  • Sajid Qureshi April 20, 2009, 6:24 am

    Hi Michael, Hello Gord!

    Thanks for your support!

    I think we should have bought 1120E/1140E instead of 1110 series. Anyways, thanks again for your advice.

    Micheal! , there is no rating for your Blog service. If it was a rating, I would like to give it 5 out of 5. Good work! ;)

    – Sajid

  • Oscar Badilla June 19, 2009, 1:56 pm

    Michael,
    I was trying to pass parameters to my 1140E phones through my MS DHCP v6 server. I created option 128 and named it as “Call Server Information” as you sugested, I created a string option type with “Nortel-i2004-A,aaa.aaa.aaa.aaa:pppp,1,5;0.0.0.0:0,0,0.”, but keep receiving parse error.
    I found MS DHCP server has some limitations, for example:

    “The Microsoft-based DHCP server usually allocates 312 bytes for DHCP options. That is more than enough for most option configurations. Some other DHCP servers and clients support option overlay, in which unused space in other standard DHCP message header fields within the DHCP packet can be overlaid to store and carry additional options. Microsoft DHCP service does not support this feature. If you attempt to use more than 312 bytes, some option settings will be lost. In that case, you should delete any unused or low-priority options”…took from “http://technet.microsoft.com/en-us/library/cc958929.aspx”

    Have you ever tried with MS DHCP server?, thks.

    -Oscar

    • Michael McNamara June 20, 2009, 8:03 am

      Hi Oscar,

      Yes it will work with Microsoft’s DHCP server. The string you described above… “Nortel-i2004-A,aaa.aaa.aaa.aaa:pppp,1,5;0.0.0.0:0,0,0.” I hope you realize that you need the IP address of your Call Server node in place of ‘aaa.aaa.aaa.aaa’ and the port you are using in place of ‘pppp’. Here’s an example that I use in my network;

      "Nortel-i2004-A,10.1.240.40:4100,1,10;10.1.240.40:4100,1,10."
      

      You’ll note that I use Alcatel-Lucent’s VitalQIP so I need to enclose the entire option within quotation marks within VitalQIP.

      Good Luck!

      • Oscar Badilla June 22, 2009, 11:02 am

        Thks Michael,

        Don´t worry about “aaa…” and “ppp…” information, I just put it to not use the real ones. It is weird, with full dhcp selected at phone keep receiving parse error, with partial dhcp and manually configured call server ip, phone connects right. I quoted string with single, double and none marks without success. Firmware version is 0625C4l.
        I don´t think it is something related with vlans nor dhcp-relay because partial dhcp is working. Seems like string is not been sent by dhcp server or phone don´t understand it.
        I will try to capture some frames to see what´s going on there…

        • Michael McNamara June 22, 2009, 5:10 pm

          Very interesting… how do you have the phone configured and what version of firmware is it running?

          If you have the opportunity a sniffer/packet trace would really help you… or if you can run your DHCP server in debug mode that will probably help too. Are you using DHCP option 191 to return the Voice VLAN to the phones?

          Cheers!

          • Oscar Badilla June 22, 2009, 7:01 pm

            Success!, this is what I did:

            1. Changed option type from string to encapsulated.
            2. Configured option 128 and 191 on both vlan´s scopes (Data and Voice).

            option 128: “Nortel-i2004-A,10.0.0.5:4100,1,5;10.0.0.5:4100,1,5.”
            option 191: “VLAN-A:2.”

            Notes:
            —–
            -both options without quote marks!
            -be sure to remove “00″ code at beginning.
            -use mnemonic “VLAN-A”, others seems not to work.

            3. I have configured switch port to “untag PVID only” (PVID pointing to Data vlan) to permit me connect later a pc on that port without problems.

            thks and rgds,

  • Andy Ivancic October 22, 2009, 12:06 pm

    Oscar, I tried your suggestion under MS DHCP v5.2 (Win2003 sp2) but there is no provision to encapsulate text (only Decimal or Hex data). When I tried option type of string I would get a DHCP parser error message.

    Telephone Terminal Type: 2002P2
    Firmware: 0604DBL

    Any other suggestions to get this going?

  • Andy Ivancic October 22, 2009, 12:33 pm

    I forgot to mention that Partial DHCP provisioning works OK.

  • Chris Johnson October 25, 2009, 6:57 pm

    An issue noted when configuring Nortel IP sets within a Cisco data environment.
    The symptoms of failure are DCHP parse error or DCHP server unreachable.
    Both option 128 and 191 are configured as DHCP standard option strings and formatted as detailed in above notes (wWindows Server 2003 so no “” marks).
    Port security and “sticky MAC” were the cause of the problems. Setting a device count of 2, one for the phone and one for the PC, each in different VLANs and setting the port to remember MAC address of devices, causes the registration process to fail. The phone originally gets its address lease from the data range. The set should then react to the VLAN-A 191 option and through away the data subnet address and be leased a voice VLAN address. However, the Cisco switch sees this activity as another device, a third device, attempting to register on the port and is barred. With port settings at 3 devices, and ceasing the sticky MAC process, the set registration works and the set ends up in the Voice VLAN.
    If it is imperative forthe port security to be set at 2 devices and sticky MAC to be implimented, then the following should be observed.
    Remove option 191 in DHCP and when setting up the phone, manually enter the 802.1Q Voive VLAN option by inserting the tag number. The set will then be directly steared to the correct DHCP scope and attain a Voice subnet IP address lease.

    • Michael McNamara October 27, 2009, 9:00 am

      Hi Chris,

      This might be acceptable for a handful of phones but when you start approaching the hundreds or thousands of phones, hard coding the phones for a specific VLAN ID is sure to be problematic down the road. Have you reviewed the configuration of the Cisco 3750E with LLDP?

      As a side note the majority of people with “DHCP Parse Error” have both DHCP options configured with-in the same scope. Only option 191 is required in the data VLAN scope while option 128 is required in the voice VLAN scope.

      Cheers!

      • Chris Johnson October 27, 2009, 10:08 am

        I completely agree with you that anything to minimize the human interaction with deployment is a must. The difficulty comes with the end users IT department. I am the voice guy, he is the data guy, it’s his network, his policies his job on the line if he upsets his manager.. the voice guy has to move. The voice guy isnt allowed to touch or interfere with the customers network configurations, but is expected to get the phone working regardless… and at no additional capital cost. This customer, as with all public school environments, wants the port locked down. Sticky MAC and 2 devices must stay, else the Nortel solution will be thrown away. At the same time, the Cisco educated IT people want the phones to deploy and work like Cisco… (lets give them analog POTS then) :).
        with the port device count at 2 and sticky MAC on, there is no way this side of armagedon, that any Nortel IP set will register with a full DHCP and VLAN = Auto setting. The set being sent to the Voice VLAN (191) will not function, the port already has two devices registering. Sticky MAC makes it doubly worse, because the MAC address of the set is already discovered and associated with the access VLAN. LLDP doesnt make it work ither, as long as the port settings are as stated. Compromise has to occure, two extra key presses on initail set deployment, to manually set the VLAN tag ID, or disable sticky MAC and increase the port device count to 3. I have had access to a lot of interoperability documents and this topic is not covered. I doubt Cisco will ever approuch it as the attitude there is “you fit in with us”. Customers cant afford to pay for a study trial. So the fact we have any (Nortel) documents at all is a bonus. Just another big vendor running roughshod over customer choices…
        There is as yet, no such thing as a zero touch IP set deployment… thats why students get to make a few bucks on big deployments.
        This customers issues are for the future. i.e. when and if the set user moves to another physical location in a different logical subnet and wants to keep the phone and the number.
        The reason the post ended up here? Because unfortunately there are not enough event error strings available to split the symptoms into smaller groupings when a set deployment fails. DHCP parse error.. DHCP server unreachable… how may causes? This is just another one and the only current solution choice availabale given the scenario.

        By the way.. I wish all blogs and forums were as informative as yours.. keep up the very good work

  • Gord October 26, 2009, 8:27 am

    Andy – DBL is also about 7 releases behind what is available now for the Phase 2 2004.

  • Chris Johnson October 26, 2009, 1:36 pm

    DCN , Unistim 3.4, is the current level at GA. Nortel has just retired a lot of the not so older versions. Like the sets I am having delivered come with D9B loaded which is no longer available to download, let alone be supported. Dont get caught out.

  • Simon November 18, 2009, 11:46 am

    Hi Michael,

    We are running Nortel CS1KE with Succession 5 and have mixed phones of 2004P2, 1140 and 1150.

    We are using full DHCP to configure all of the phones and this has been working fine for several months. The last couple of days however, the vast majority of phones are dropping the connection and displaying ‘DHCP Server unreachable’

    We can ping the phone from the DHCP server fine. Rebooting the phones does not make any difference, nor does deleting the lease from DHCP server or rebooting the DHCP server. After an hour or so the phones come back of their own accord??

    I have never seen anything like this, and it does not effect all of the phones.

    Has anyone experienced this, or any ideas on a possible cause/solution?

    Many thanks, Simon

Leave a Comment