Re: [DNSOP] Request for Comments on I-D about IoT DNS Name Autoconf

["Mr. Jaehoon Paul Jeong" <[e-mail]>]

Stuart,
by mistake, my incomplete email was sent.
Here is the complete email.

On Thu, Nov 5, 2015 at 2:51 PM, Stuart Cheshire <[e-mail]> wrote:

> On 3 Nov 2015, at 01:51, Mr. Jaehoon Paul Jeong <[e-mail]>
> wrote:
>
> > Hi 6man, 6lo and dnsop folks,
> >
> > There will be a talk about IoT DNS Name Autoconfiguration
> > in 6man WG's morning session tomorrow, 11/4/2015.
> >
> > Title: DNS Name Autoconfiguration for Internet of Things Devices
> > https://tools.ietf.org/html/draft-jeong-6man-iot-dns-autoconf-00
>
> This was not actually discussed in 6man, 6lo, or dnsop, so I’ll make some
> comments here.
>
> It’s hard to know where to start.
>
> Your document confuses device discovery with service discovery. What a
> device *is* tells you virtually nothing about what it *does*. The “device
> category” of my computer being “laptop” or “tablet” tells you *nothing*
> about what services it offers.
>
> Your document assumes that every search domain your tablet encounters (
> starbucks.com, narita-airport.co.jp, meeting.ietf.org, comcast.com) will
> allow your tablet to create global records in that domain. Clearly this is
> nonsense.
>
>    >> Device model (denoted as device_model) in my proposed DNS name
format can let an IoT device refer to
        the specification of another device's functions, assuming that such
a device model's specification is available
        publicly. Of course, we can use service discovery for device
functions through dnssd.
        This service discovery is the next step in my draft.
        For example, for a given Samsung's refrigerator model, such as
RF4287HARS
        (28 cu. ft. French Door Refrigerator Stainless Steel), we can know
the functions with the specification.
        See Samsung's refrigerators:
http://www.samsung.com/us/support/appliances/refrigerators

Having put global address records into starbucks.com, your document assumes
> then assumes that starbucks.com will then allow you do to a zone transfer
> to fetch the entire zone to discover the names of all the other address
> records in starbucks.com. Clearly this is nonsense too.
>
   >> In Section 7 (DNS Name Management for Mobile IoT Devices),

https://tools.ietf.org/html/draft-jeong-6man-iot-dns-autoconf-00#section-7,
        I discuss the mobility issue of an IoT device to receive multiple
search domains.
        Whenever the IoT device recognizes the movement into another
subnet, it can delete its DNS names
        by DNS dynamic update.
        In reality, in public areas (e.g., starbucks and narita airport),
we can disable the automatic DNS name
        registration in routers because the registration in the public
areas makes some privacy issues.
        We can enable such automatic registration in our managed networks,
such as home network, office network,
        smart grid, and factory network.
        Thus, we can prevent your concern for global DNS names from
happening.

>
> Your document proposes global address records with names with this form:
>
> unique_id.device_model.device_category.mic_loc.mac_loc.domain_name.
>
> For example:
>
>
> jkadjkhdsafhjlsadfjklkljdgajknsadf.Sungkyunkwan-1234.cleaning-robot.right-upper-corner.living-room.comcast.com
> .
>
> The host name of the cleaning robot keeps changing as it moves around the
> room, requiring continual updates and continual zone transfers to keep
> track of the name as it changes. Clearly this is infeasible.
>
   >> To prevent the frequent update of a mobile IoT device's DNS name, the
IoT device can update its DNS name
        in a reasonable time interval. This is a further discussion issue.
        My point is that the physical location in the DNS name can help
users easily track the position of an IoT device.

>
> I would, however, love to get one of these new flying cleaning robots,
> which can be located (as it was in your example), “in the right-upper
> corner of a living room.”
>
   >> Me too :-)

    Thanks.

    Paul


>
> Stuart Cheshire
>
>


-- 
===========================
Mr. Jaehoon (Paul) Jeong, Ph.D.
Assistant Professor
Department of Software
Sungkyunkwan University
Office: [phone]
Email: [e-mail], [e-mail]
Personal Homepage: http://cpslab.skku.edu/people-jaehoon-jeong.php