Re: [dhcwg] RFC 3925 Data Length Placement

[Bernie Volz <[e-mail]>]

Hi:

While I agree that there is an overall length for the vendor-class data, it must be formatted as per RFC:

The vendor-class-data comprises a series of separate items, each of
   which describes some characteristic of the client's hardware
   configuration or capabilities.  Examples of vendor-class-data
   instances might include the version of the operating system the
   client is running or the amount of memory installed on the client.

   Each instance of the vendor-class-data is formatted as follows:

                        1 1 1 1 1 1
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |   data-len    |               |
   +-+-+-+-+-+-+-+-+  opaque-data  |
   /                               /
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   The data-len is one octet long and specifies the length of the opaque
   vendor class data in network byte order.

There’s also a MUST for how the Vendor Information option is formatted … so it is not up to the vendor.

The reason for an overall length was that it made the server’s processing when skipping a specific vendor’s data easier. (And also avoided severe problems if a vendor didn’t follow the rules.)

So, if there is one vendor class of “abcd”, it is formatted as octet of 5, 4, and then “abcd”. If there are two vendor classes “abcd” and “efgh”, it is formatted as octet of 10, 4, abcd, 4, efgh. The 4 octet followed by the 4 bytes represent the diagram above for each piece of opaque-data. The overall length of 5 or 10 is the data-len1 (or 2, …).

- Bernie Volz

On Apr 4, 2023, at 5:11 PM, Ted Lemon <[e-mail]> wrote:



There is a length per enterprise and a length for the total option. The format of the per-enterprise data is up to the enterprise to define.

On Tue, Apr 4, 2023 at 4:04 PM Timothy Winters <[e-mail]> wrote:

Hello,

As part of CPEs implementing https://usp.technology/specification/03-index-discovery.html#sec:using-dhcp" target="_blank" rel="nofollow">DHCP Discovery for USP, we had several implementations recently ask about the packet format of Option 124.  In particular the data-len fields documented in RFC 3925, Section 3.  There is confusion if there is a single data length following each enterprise-number for the VI Vendor Class or if the vendor-class-data option has a data length as well.  

Single Data Len

8 (data-len1)

   abcd (first class)

   defg (second class)

OR 

Multiple Data Len

8 (data-len1)

  3 abc (first class)

  3 def (second class)

Regards,

Tim

_______________________________________________
dhcwg mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/dhcwg" rel="noreferrer nofollow" target="_blank">https://www.ietf.org/mailman/listinfo/dhcwg

_______________________________________________
dhcwg mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/dhcwg