#35: Last call comments: editorial
Comment (by [email protected]):
Replying to [ticket:35 shwethab@Š]:
'''Chonggang:'''
· pp. 3, 2nd paragraph, ³Šcontrol operations such industrial
automationŠ² ---> ³³Šcontrol operations such as industrial automationŠ²
· pp.3, 3rd paragraph, ³ŠtimeSlotted Channel Hopping Š² --->
³ŠTime Slotted Channel HoppingŠ²
· pp.3, 4th paragraph, the fulling spelling of TSN was givning
in the 2nd paragraph. Then, please put the acronym TSN in the 2nd
paragraph
· pp.3, both ³deterministic² and ³Deterministic² are used in a
few places. Better use one form if they mean the same thing throughout
the
draft.
· pp.4, 2nd paragraph ³Route Computation may be achieved in a
centralized fashion by a Path Computation Element (PCE), in a distributed
fashion using the Routing Protocol for Low Power and Lossy Networks (RPL)
[RFC6550], or in a mixed mode²,
o Is RPL claim to be a distributed route computation? I believe in
the
non-storing mode, the root calculates the route and use source routing to
route the packet to the destination.
· pp.4, last two paragraphs from the bottom and a few other
place in the draft, ³Šneighbor DiscoveryŠ² ---> ³Š Neighbor DiscoveryŠ²
· pp. 5, section 3, 1st paragraph, ³Some aspects of this
architecture derive from existing industrial standards for Process
Control
by its focus on Deterministic Networking, in particular with the use of
the IEEE802.15.4e [IEEE802154e] TSCH MAC and a centralized PCE.²
o What are the ³existing industrial standard²? WirelessHart? Certain
references would help.
· pp.6, the title of Figure could be more complete by saying
something like ³Figure 1, Basic Configuration of 6TiSCH Network²
· pp. 6, the paragraph under Fig. 1, ³neighbor Devices² --->
³neighbor devices²
· pp.6, 2nd paragraph from the bottom, ³LLN Border Router
(LBR)²
were mentioned twice in one sentence.
· pp.8, Figure 3, suggest changing its title to ³6TiSCH
Protocol
Stack²
· pp.8, Figure 3, ³6LoWPAN HC² shows between IPv6 and 6top. It
is correct but it could imply that only HC is used there which is not
true. Suggest changing 6LoWPAN HC to something like ³6LoWPAN (e.g.
adaptation, header compression)²
· pp.9, the 4th paragraph, ³join process² was introduced all of
a sudden. Wonder if it could read more smoothly by briefly mentioning
these concepts (e.g. neighbor discovery, join process, track reservation,
6top, packet forwarding, etc.) together and their relationship from
architectural perspective somewhere in Section 4 (e.g. at the
beginning?).
· pp.9, 3rd paragraph, ³TEAS protocol will be required to
expose
the device capabilities and the network peerings to the PCE, and a
protocol such as a lightweight PCEP or an adaptation of CCAMP G-MPLS
formats and procedures will be used to publish the tracks, computed by
the
PCE, to the devices.²
· Can you put the reference about ³TEAS² and ³CCAMP²?
· pp.10, 2nd paragraph, ³should be portable only other LLN link
types². It reads a little awkward and maybe some words missing around
³only²
· pp.10, 2nd paragraph, ³A new version of the standard is
expected in 2015². What¹s ³the standard² refer to?
· pp.10, 4th paragraph, do you want to add a reference for
ISA100.20?
· pp.16, 4th paragraph, ³motes² was used. Can we change it to
devices or nodes to make terms more consistent?
· pp.17, 2nd paragraph, ³but also allows an upper layer like
RPL.² This sentence seems not finished
· pp.18, 2nd paragraph, ³that is is not capable of². One ³is²
should be removed.
· pp.19, the last paragraph, ³an higher layer² ---> ³a higher
layer²
· pp.19, the last paragraph, ³DSCP can therefore be used².
Suggest giving the full spelling and reference to DSCP or removing this
sentence.
· pp.19, 5th paragraph, ³A 6TISCH Instance is associated to one
slotFrame.²,
· Wonder what is a 6TiSCH Instance? Why associate to only ³one²
Slotframe? In previous paragraph, it mentions ³Multiple slotFrames can
coexist in a node schedule².
· pp.21, section 9, both ³Static Scheduling² and ³Minimal
Static
Scheduling² are used. Are they the same?
· pp.22, 1st paragraph, ³This scheduled can be² ---> ³This
schedule can be²
· pp.24, section 10.1, 1st paragraph, ³A bundle of cells set to
receive (RX-cells) is uniquely paired to a bundle of cells that are set
to
transmit (TX-cells), representing a layer-2 forwarding state that can be
used regardless of the network layer protocol.²
· Is a Track uni-directional or bi-directional? Is there any
ACK
for the transmission?
· pp.27, Figure 6, suggest adding what is the value ³set dmac
to² and ³restore dmac²
· pp.27, section 10.1.3, 1st paragraph, ³If the tunnel egress
point does not have a MAC address that matches the configuration, the
Track installation fails.²,
· Wonder if it is ingress point or egress point that installs
the Track
· pp.30, 1st paragraph, ³Centralized routes can for example
computed². The word ³be² was missing.
· pp.30, section 11, it will be helpful if the difference
between routing discussed in this section and forwarding discussed in
section 10 could be clarified a little bit
----
Anand:
There is just one observation which I could not resist mentioning. It
has to do with
the coupling of real-time application demands with the rest of the
architecture. There could be
application specific "time critical and sudden" events that might call
for on-demand resource allocation.
While we can delegate such an event handling to NME and PCE, perhaps,
bringing in an application awareness
to the overall architecture might be useful so that we are not leaving
out this possible scenario.
----
Maria Rita:
pag 8 LLN odes -> LLN nodes
pag. 19 6TISCH -> 6TiSCH
pag 23 as started -> has started
pag 19 in the definition of CDU matrix, I would specify height and
width, and after the timeslot duration (in the current version timeslot
duration is in the middle, and it makes a bit hard to read and clearly
understand the CDU matrix structure)
pag 7 are we sure we want to mention in the next round of the document
we will have evaluated PANA applicability? we may need/want to publish
the
second part of the architecture before that, we don¹t know yet. So, we
could skip that, if you agree.
----
Rouhollah:
Page 7 first paragraph. LLN odes
Page 9 end of third paragraph. foster
1-Do registration phase may work properly when network builds up, if
when RPL tree grows up and extends like a complete tree (I mean in
distributed manner). Maybe some additional controls needed later?
2-Do you think DICE(figure 3. page 7) should be defined in the
Terminology draft.
Typos:
Page 21. scheduled
----
Guillaume:
1) The term reallocation/relocation of cells :
8.1 p16 a cell reallocation
differs from
9.2 p22 the relocation of a soft cell / relocation is done
=> I am not sure if there has been a discussion on this term, but I
vote
for reallocation.
2) 8.2 p16 a set of quality metrics (RSSI, LQI)
I think you cannot be specific to just two metrics (RSSI, LQI).
The words "a set of" allows to define more metrics. And I think it is
fine.
=> what do you think of : "a set of quality metrics (e.g. RSSI, LQI,
etc.) " ?
3) 8.2 p16 and used to compute a Rank Increment
The statistics of 6top may have other usages than incrementing the
ranks, for upper layers, and for resource management.
=> what about : "and used for instance to compute" ?
4) 10.2 p28 : a degree of flow control base on => I think it should be
"based on".
5) (Already pointed out) 11 p 30 Centralized routes con for example
computed => be computed
----
Jonathan Simon:
13 - Sending link-layer frames in the clear in the initial stage of
joining is not providing any benefit. We should always use
authentication,
even if the key is not secret, as it provides the ability to reject
similar frames from other 802.15.4-based protocols. It also isn¹t
necessary to discuss such a detail here.
13.1 -
* "Triage" - So the JCE decides which nodes are more important and
assigns resources to them first? How? Note this term is not used in
draft-richardson-6tisch-security-architecture-02.
* "arbitrage" should be ³arbitrate²
Other than that, it seem to be capturing the overall spirit of the
security architecture and highlights the open areas of security
discussion, e.g. that PANA is an open issue.
Couple minor points:
1) Introduction and 3) Application and Goals sections need review. They
kind of wander all over the place.
5.1) What does "volume" mean in this context? Is this referring to
other
as yet defined documents, or later revisions of this document?
----
Rene:
One note: the second para of Clause 13 seems out of place and should be
removed.
----
Answer
------
I addressed all the comments doing one commit per; and answer on the ML.
The list of commits is in https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/all , each with a diff.
I made the following changes; based on Michael's comments
Pascal Thubert ae7ea78 07 2015-03-10 https://bitbucket.org/6tisch
/draft-ietf-6tisch-
architecture/commits/ae7ea789ac066cb6b31fcdf9e6e8fdc2f6cf667e
Pascal Thubert 62b2606 06 final 2015-03-09
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/62b2606b48d14cb83f4d1d87c600c3f333cc8f66
Pascal Thubert 2b29626 trailing whitespace removal 2015-02-24
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/2b296269f964eae8a7d47216ed90cf969c405a9d
Pascal Thubert 28c6316 Maria-Rita's review 2015-02-24
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/28c6316289ff06e21adc340c962acca94277de22
Discussion
----------
typos to be fixed:
pag 8 LLN odes -> LLN nodes
pag. 19 6TISCH -> 6TiSCH
pag 23 as started -> has started
-> all fixed
Moreover,
page 19 in the definition of CDU matrix, I would specify height and
width,
and after the timeslot duration (in the current version timeslot duration
is in the middle, and it makes a bit hard to read and clearly understand
the CDU matrix structure)
-> text now reads:
In order to describe that formatting of time and frequencies,
the
6TiSCH architecture defines a global concept that is called a
Channel
Distribution and Usage (CDU) matrix; a CDU matrix is a matrix of
cells with an height equal to the number of available channels
(indexed by ChannelOffsets) and a width (in timeSlots) that is
the
period of the network scheduling operation (indexed by
slotOffsets) for
that CDU matrix. The size of a cell is a timeSlot duration, and
values of 10 to 15 milliseconds are typical in 802.15.4e TSCH
to
accommodate for the transmission of a frame and an ack,
including
the
security validation on the receive side which may take up to a
few
milliseconds on some device architecture.
pag 7 are we sure we want to mention in the next round of the document we
will have evaluated PANA applicability? we may need/want to publish the
second part of the architecture before that, we don¹t know yet. So, we
could skip that, if you agree.
-> I¹d like to leave things open. What about:
Future work on 6TiSCH security and will examine in deeper detail
how
SACM,
ACE, DICE, IEEE802.15.9 and eventually PANA and 802.1x may apply to
6TiSCH
networks to perform Authentication and Authorization, to secure
transactions
end-to-end, and to maintain the security posture of a device over
its lifetime.
The result of that work will be described in a subsequent volume of
this
architecture.
Pascal Thubert 0e07e72 René Struik's review 2015-02-24
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/0e07e721ce096cbd02f93c5cfbf9e58e576ac124
Comments:
---------
I removed that section, and added an appendix that lists the 6TiSCH
personal submissions that may impact the next volumes.
The new text reads as follows
Appendix A. Personal submissions relevant to the next volumes
This volume only covers a portion of the total work that is needed to
cover the full 6TiSCH architecture. Missing portions include
Deterministic Networking with Track Forwarding, Dynamic Scheduling,
and Security.
[I-D.richardson-6tisch-security-architecture] elaborates on the
potential use of 802.1AR certificates, and some options for the join
process are presented in more details.
[I-D.struik-6tisch-security-architecture-elements"]
describes 6TiSCH security architectural elements with
high level requirements and the security framework that are relevant
for the design of the 6TiSCH security solution.
[I-D.dujovne-6tisch-on-the-fly] discusses the use of the 6top
sublayer [I-D.wang-6tisch-6top-sublayer] to adapt dynamically the
number of cells between a RPL parent and a child to the needs of the
actual traffic.
Pascal Thubert c34d7eb ChongGang's review 2015-02-24
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/c34d7ebf8abb67a51fb94717480fcc5cc931c572
Pascal Thubert 1a72a0f updated 2015-02-24 https://bitbucket.org/6tisch
/draft-ietf-6tisch-
architecture/commits/1a72a0f578371be194832d6525c1e502e0055b92
Comments
--------
Please see below for details:
I have reviewed the architecture draft (-05) and support to publish it.
A few minor comments I had are listed below.
€ pp. 3, 2nd paragraph, ³Šcontrol operations such industrial
automationŠ² ---> ³³Šcontrol operations such as industrial automationŠ²
-> done
€ pp.3, 3rd paragraph, ³ŠtimeSlotted Channel Hopping Š² --->
³ŠTime Slotted Channel HoppingŠ²
-> Actually I think the group settled for TimeSlotted, I migrated all to
that
€ pp.3, 4th paragraph, the fulling spelling of TSN was givning in
the 2nd paragraph. Then, please put the acronym TSN in the 2nd paragraph
-> It was but I changed the format to be more classical
€ pp.3, both ³deterministic² and ³Deterministic² are used in a
few
places. Better use one form if they mean the same thing throughout the
draft.
-> Done, I used the uppercase because our Deterministic is not the one
from math
€ pp.4, 2nd paragraph ³Route Computation may be achieved in a
centralized fashion by a Path Computation Element (PCE), in a distributed
fashion using the Routing Protocol for Low Power and Lossy Networks (RPL)
[RFC6550], or in a mixed mode²,
o Is RPL claim to be a distributed route computation? I believe in the
non-storing mode, the root calculates the route and use source routing to
route the packet to the destination.
-> I see what you are saying. I agree that storing mode centralized the
computation of routes down but that¹s based on a parent selection that is
still distributed. I do not think that being very specific like ³RPL is
mostly distributed but then there can be a bit of centralized
computation²
will help a lot the understanding of the section. I¹d rather leave it at
that.
€ pp.4, last two paragraphs from the bottom and a few other place
in the draft, ³Šneighbor DiscoveryŠ² ---> ³Š Neighbor DiscoveryŠ²
-> sure, many occurrences fixed
€ pp. 5, section 3, 1st paragraph, ³Some aspects of this
architecture derive from existing industrial standards for Process
Control
by its focus on Deterministic Networking, in particular with the use of
the IEEE802.15.4e [IEEE802154e] TSCH MAC and a centralized PCE.²
o What are the ³existing industrial standard²? WirelessHart? Certain
references would help.
-> OK, added
€ pp.6, the title of Figure could be more complete by saying
something like ³Figure 1, Basic Configuration of 6TiSCH Network²
-> OK
€ pp. 6, the paragraph under Fig. 1, ³neighbor Devices² --->
³neighbor devices²
-> OK
€ pp.6, 2nd paragraph from the bottom, ³LLN Border Router (LBR)²
were mentioned twice in one sentence.
->reworded
€ pp.8, Figure 3, suggest changing its title to ³6TiSCH Protocol
Stack²
-> Yes, I added envisioned since all is not fully cast in stone
€ pp.8, Figure 3, ³6LoWPAN HC² shows between IPv6 and 6top. It is
correct but it could imply that only HC is used there which is not
true.
Suggest changing 6LoWPAN HC to something like ³6LoWPAN (e.g. adaptation,
header compression)²
-> added
€ pp.9, the 4th paragraph, ³join process² was introduced all of a
sudden. Wonder if it could read more smoothly by briefly mentioning these
concepts (e.g. neighbor discovery, join process, track reservation, 6top,
packet forwarding, etc.) together and their relationship from
architectural perspective somewhere in Section 4 (e.g. at the
beginning?).
-> will be hard without duplicating a lot; let me think of that one. In
any fashion, I can clarify join there. I propose
Security is often handled at Layer-2 and Layer 4. Authentication
during the process on joining or re-joining the network
is discussed in <xref target="sec"/> and the
applicability of existing protocols such as
the <xref target="RFC5191">
Protocol for Carrying Authentication for Network access
(PANA)</xref>
will be studied in a next volume of this document.
€ pp.9, 3rd paragraph, ³TEAS protocol will be required to expose
the device capabilities and the network peerings to the PCE, and a
protocol such as a lightweight PCEP or an adaptation of CCAMP G-MPLS
formats and procedures will be used to publish the tracks, computed by
the
PCE, to the devices.²
€ Can you put the reference about ³TEAS² and ³CCAMP²?
-> added, text looks like this now:
The work on centralized track computation is deferred to a
subsequent volume
of the architecture. The Path Computation Element (PCE) is
certainly
the core component of that architecture. Around the PCE, a
protocol
such as an extension to a TEAS <xref target="TEAS"/> protocol
(maybe running over CoAP as illustrated) will be required to
expose the
device capabilities and the network peers to the PCE, and a
protocol
such as a lightweight PCEP or an adaptation of CCAMP <xref
target="CCAMP"/>
G-MPLS formats and procedures will be used to publish the
tracks,
computed by the PCE, to the devices (maybe in a fashion similar
to RSVP-TE).
€ pp.10, 2nd paragraph, ³should be portable only other LLN link
types². It reads a little awkward and maybe some words missing around
³only²
->sure! The text now reads
The current charter positions 6TiSCH on IEEE802.15.4 only.
Though most of the design should be portable on other link types,
6TiSCH has a strong dependency on 802.15.4 and its evolution.
€ pp.10, 2nd paragraph, ³A new version of the standard is
expected
in 2015². What¹s ³the standard² refer to?
-> clarified as follows
The current charter positions 6TiSCH on IEEE802.15.4 only.
Though most of the design should be portable on other link types,
6TiSCH has a strong dependency on IEEE802.15.4 and its evolution.
A new version of the IEEE802.15.4 standard is expected in 2015.
€ pp.10, 4th paragraph, do you want to add a reference for
ISA100.20?
-> there¹s no good link. I added a link to ISA100, and explained what CNM
is.
ISA100 <xref target="ISA100"/> Common Network Management (CNM) is
another
external work of interest for 6TiSCH. The group, referred to as
ISA100.20,
defines a Common Network Management framework that should enable
the
management of resources that are controlled by heterogeneous
protocols
such as ISA100.11a <xref target="ISA100.11a"/>, WirelessHART
<xref target="WirelessHART"/>, and 6TiSCH. Interestingly, the
establishment of 6TiSCH Deterministic paths, called tracks,
are also in scope, and ISA100.20 is working on requirements for
DetNet.
€ pp.16, 4th paragraph, ³motes² was used. Can we change it to
devices or nodes to make terms more consistent?
-> moved to devices, but added the following
on behalf of the wireless devices (also called motes),
€ pp.17, 2nd paragraph, ³but also allows an upper layer like
RPL.²
This sentence seems not finished
-> OK, the sentence now looks like
The receiver of the Enhanced Beacon MAY
be listening at the cell to get the Enhanced Beacon
([IEEE802154e]).
6top takes this way to establish broadcast channel, which not
only
allows TSCH to broadcast Enhanced Beacons, but also allows
protocol
exchanges by an upper layer such as RPL.
€ pp.18, 2nd paragraph, ³that is is not capable of². One ³is²
should be removed.
-> ack
€ pp.19, the last paragraph, ³an higher layer² ---> ³a higher
layer²
-> ack
€ pp.19, the last paragraph, ³DSCP can therefore be used².
Suggest
giving the full spelling and reference to DSCP or removing this sentence.
-> I used ŒDifferentiated Services [RFC2474]¹ instead
€ pp.19, 5th paragraph, ³A 6TISCH Instance is associated to one
slotFrame.²,
-> that¹s really a 6top internal. Much too detailed, I removed the
sentence.
€ Wonder what is a 6TiSCH Instance? Why associate to only ³one²
Slotframe? In previous paragraph, it mentions ³Multiple slotFrames can
coexist in a node schedule².
-> same, that is an internal construct
€ pp.21, section 9, both ³Static Scheduling² and ³Minimal Static
Scheduling² are used. Are they the same?
-> removed minimal. Intention was to say that this is the minimal support
€ pp.22, 1st paragraph, ³This scheduled can be² ---> ³This
schedule can be²
-> gone
€ pp.24, section 10.1, 1st paragraph, ³A bundle of cells set to
receive (RX-cells) is uniquely paired to a bundle of cells that are set
to
transmit (TX-cells), representing a layer-2 forwarding state that can be
used regardless of the network layer protocol.²
€ Is a Track uni-directional or bi-directional? Is there any ACK
for the transmission?
-> I added
A Track is a unidirectional path between a source and a
destination.
In a Track cell, the normal operation of IEEE802.15.4e
Automatic Repeat-reQuest (ARQ) usually happens, though the
acknowledgment may be omitted in some cases, for instance if
there
is no scheduled cell for a retry.
€ pp.27, Figure 6, suggest adding what is the value ³set dmac to²
and ³restore dmac²
-> done
€ pp.27, section 10.1.3, 1st paragraph, ³If the tunnel egress
point does not have a MAC address that matches the configuration, the
Track installation fails.²,
€ Wonder if it is ingress point or egress point that installs the
Track
I think the text is correct. The configuration of the tunnel is yet TBD
by
this group (RSVP-TE?), but the egress must be able to deliver a packet to
the MAC address that is at the end of the tunnel.
€ pp.30, 1st paragraph, ³Centralized routes can for example
computed². The word ³be² was missing.
fixed
€ pp.30, section 11, it will be helpful if the difference between
routing discussed in this section and forwarding discussed in section 10
could be clarified a little bit
Added
By forwarding, this specification means the per-packet operation
that
allows to deliver a packet to a next hop or an upper layer in
this node.
Forwarding is based on pre-existing state that was installed as
a
result of a routing computation.
Pascal Thubert cca7835 Rouhollah's comments 2015-02-23
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/cca78350db5d5ad3a8b031944e44ce15972453ae
Comments
---------
1-Do registration phase may work properly when network builds up, if when
RPL tree grows up and extends like a complete tree (I mean in
distributed
manner). Maybe some additional controls needed later?
Unsure what you really mean here. There can be issues with security and
the join process work is looking at that. Then there¹s RPL use of slotted
aloha slots. Then there¹s the resource allocation problem (chunks between
parents and cells in a parent¹s chunk between children). We¹ll recharter
for all these things but the first volume of the architecture does not
cover that.
2-Do you think DICE(figure 3. page 7) should be defined in the
Terminology
draft.
Well, that¹s a WG now, unsure what the standard name will be. I can add
the following text:
³ DTLS In Constrained Environments (DICE) is at the time of this
writing the probable way LLN nodes will provide end-to-end
security for
UDP/CoAP packets.²
Typos:
Page 21. scheduled
Ack
On Mon, Feb 16, 2015 at 8:10 PM, Rouhollah Nabati(Google)
<[email protected]> wrote:
Hello All,
Page 7 first paragraph. LLN odes
ack
Page 9 end of third paragraph. foster
I missed the issue there?
I agree with all Guillaume¹s suggested edits. Changes:
Pascal Thubert 5c2a38f Guillaume's edits 2015-02-23
https://bitbucket.org/6tisch/draft-ietf-6tisch-
architecture/commits/5c2a38ffd0b74695b26551a0ca5b90fd0ab3d7fb
--
-------------------------+------------------------------------------------
-
Reporter: | Owner: draft-ietf-6tisch-
[email protected] | [email protected]
Type: defect | Status: new
Priority: minor | Milestone:
Component: | Version:
architecture | Resolution:
Severity: In WG Last |
Call |
Keywords: |
-------------------------+------------------------------------------------
-
Ticket URL:
<http://trac.tools.ietf.org/wg/6tisch/trac/ticket/35#comment:1>
6tisch <http://tools.ietf.org/6tisch/>
_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch
