[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[DNA] RE: draft-ietf-dna-link-information-03




CC: DNA WG ML.

Thanks Jari for your review and comments.

Please see my inlined response below.

> -----Original Message-----
> From: Jari Arkko [mailto:jari.arkko@piuha.net]
> Sent: Monday, May 22, 2006 10:29 AM
> To: alper.yegin@yegin.org
> Cc: Greg Daley; Suresh.Krishnan@ericsson.com; gregdaley@ieee.org
> Subject: Re: draft-ietf-dna-link-information-03
> 
> Here are my initial AD review comments. I'm not
> completely done yet, because I want to re-read
> the IAB link indications document to see if there's
> something else that should be taken into account.
> 
> Btw, Bernard reviewed this document at some
> point. Where is that review, and did you implement
> all of his suggestions?

We had done done per his earlier review. I'll respond to his new review
comments as well.


> Substantial:
> 
> >    Two basic link-layer events are considered potentially useful to DNA
> >    process: link up and link down.  Both of these events are
> >    deterministic, and their notifications are provided to IP-layer after
> >    the events successfully conclude.
> 
> I wonder how deterministic these events are in a wireless
> network.

I think it is best if we hadn't used that term. I'd rewrite:

Two basic link-layer events are considered potentially useful to DNA
process: link up and link down.  Associated notifications can be
provided to the IP-layer after the events conclude successfully.

 
> >    A GPRS MT that wants to establish IP-level connections MUST first
> >    perform a GPRS attach to the SGSN.  This MUST be followed by a
> 
> This reads like normative language for the implementation of
> GPRS L2 operations. I would suggest that such language be
> left to the SDOs in question. (May apply to other parts of
> the document, too. Go over the MUSTs and SHOULDs and check
> that they apply only to the up/down indications, not the
> rest of the L2 operation.)

I'll take care of them all.


> 
> >    A GPRS MT that wants to establish IP-level connections MUST first
> >    perform a GPRS attach to the SGSN.  This MUST be followed by a
> >    request to the GPRS network to settle the necessary soft state
> >    mechanism (GPRS tunneling protocol) between its serving SGSN and the
> >    GGSN.  The soft state maintained between the MT, the SGSN and the
> >    GGSN is called a PDP Context.  It is used for guaranteeing a
> >    negotiated quality of service for the IP flows exchanged between the
> >    GPRS MT and an external Packet Data Network such as Internet.
> 
> Hmm... there's more to a PDP context than QoS state. I would simply
> replace the entire text above with:
> 
> "A GPRS MT that wants to establish IP connectivity establishes
> first a connection to the GPRS network and one or more PDP
> Context associations between the MT and the GGSN."

Sounds reasonable to me. Unless Eric, the author of that text, has any
comments, I'd go with your suggestion.


> 
> >    An MT has the possibility to establish a secondary PDP Context while
> >    re-using the IP configuration acquired from a previously established
> >    and active PDP Context.  Establishment of a secondary PDP Context
> >    does not provide additional information to IP-layer.  Such a second
> >    PDP Context would basically have a different QoS profile so that a
> >    different type of application can be served.  In that case,
> >    activation of the secondary PDP Context MUST NOT generate another
> >    link up event notification.  However, a secondary PDP Context
> >    establishment that triggers a new IP configuration is to be treated
> >    from the IP layer as indicated above.
> 
> Rewrite as
> 
>  "An MT has the possibility to establish a secondary PDP Context while
>    re-using the IP configuration acquired from a previously established
>    and active PDP Context.  Such a secondary PDP Context
>    does not provide additional information to IP-layer and only
>    allows another QoS profile to be used.  The activiation of a such
>    secondary PDP Context MUST NOT generate another
>    link up event notification.  However, other additional PDP Context
>    activiations are to be treated as indicated earlier."

My response is same as the previous one.

> 
> >    In a WiFi network that supports Robust Secure Network (RSN
> >    [IEEE-802.11i]), successful completion of 4-way handshake between the
> >    STA and AP commences the availability of IP service.
> 
> Presumably mere support is not the key factor here -- the question
> is whether the RSN operation has been turned on or not; my RSN-capable
> network works fine even without 4-way handshake.

I'd rewrite this as:

In a WiFi network that uses Robust Secure Network (RSN
[IEEE-802.11i]), successful completion of 4-way handshake between the
STA and AP commences the availability of IP service.


> 
> > 2.3  IEEE 802.11/WiFi
> 
> Does the document provide any guidance on what the link up/down
> events should be for preauthentication or fast handoff (802.11r)
> situations? Or is it too early to specify these?

It does not. I am not inclined to get into those areas, unless there is
popular demand :-)

> > 4.  Security Considerations
> >
> >    A faked link-layer event notification can be used to launch a
> >    denial-of service attack on the node and the associated network.
> >    Secure generation and delivery of these notifications MUST be
> >    ensured.  This is a subject for link-layer and network stack designs
> >    and therefore it is outside the scope of this document.
> 
> This paragraph appears to define a MUST requirement for
> something that is commonly not available. It would be useful
> to describe the vulnerability and discuss in turn each
> mitigating mechanism. You may also wish to take a look
> at Section 5.4,  Spoofing Network Connectivity Indications,
> in draft-ietf-mobike-protocol-08.txt.
> 
> For instance, the mitigating mechanisms include
> - link layer security (but should note that it can never fully
>   protect against DoS-type attacks, and link-down could
>   still be generated)
> - security mechanisms in the IP layer configuration,
>   including DHCP security and SEND security for ND/RD;
>   these can be used by
> 
> And yes, there's no complete protection today for this.

I think it is best if we didn't get into solving this problem in this
document. I am not sure if we can really get away with few lines of text to
suggest some solutions. A thorough solution (set) is likely to be pretty
comprehensive -- possible an I-D on its own. 

I think acknowledging the threat and declaring its solution out-of scope is
what we should be doing. 

As for the "MUST", I understand that there are currently deployed systems
that do not deploy appropriate security to handle this threat. That fact
should not take anything away from what a secure system MUST be doing. I
think that fact just makes them insecure in this sense, hence vulnerable to
spoofed link ups and downs. For that, I'm inclined to keep the MUST in
there. 



> 
> Editorial:
> 
> >    accomodate implementations that desire to indicate the link is up but
> 
> Typo.
> 
> >               specific ations - Media access control (MAC) Bridges"
> 
> Typo.
> 
> Line 355 has weird spacing: '...ss, but  the i...'
> 
> Missing Reference: [DNA4] is mentioned on line 404, but not defined
> 
> >   Certain network access technologies are capable of providing various
> >   link-layer status information to IP.
> 
> Maybe s/various// or s/various/various types of/
> 
> 
> >   layer cannot make a determination about the faith of these packets,
> 
> s/faith/fate/
> 
> >   GPRS is an enhancement to the GSM data transmission architecture and
> >   capabilities, designed to allow for packet switching in user data
> >   transmission within the GPRS network as well as for higher quality of
> >   service for the IP traffic of Mobile Terminals with external Packets
> >   Data Networks such as the Internet or corporate LANs [GPRS][GPRS-
> >   LINK].
> 
> I would replace this complicated sentence with this:
> 
> "GSM Packet Radio System (GPRS) provides packet switched data transmission
> over a cellular network."

Thanks for catching these as well.

Alper


> 
> --Jari
>