All

I offered a use case late last year.  The thread subject was: AllJoyn Root CA & 
Device Certificates
I have pasted a portion of that thread below. It is worth going into the mail 
archives and reading the related messages to get the inputs from others.

From:[email protected]<mailto:[email protected]>
 [mailto:[email protected]] On Behalf Of Markham, 
Tom

Sent: Tuesday, October 27, 2015 11:05 AM

To: 
[email protected]<mailto:[email protected]<mailto:[email protected]%3cmailto:[email protected]>>;
 
[email protected]<mailto:[email protected]<mailto:[email protected]%3cmailto:[email protected]>>;
 
[email protected]<mailto:[email protected]<mailto:[email protected]%3cmailto:[email protected]>>;
 
[email protected]<mailto:[email protected]<mailto:[email protected]%3cmailto:[email protected]>>;
 
[email protected]<mailto:[email protected]<mailto:[email protected]%3cmailto:[email protected]>>

Subject: Re: [Allseen-core] [Allseen-ssc] AllJoyn Root CA & Device Certificates

First , I apologize that my other responsibilities prevent me from spending as 
much time working Allseen/AllJoyn as it deserves and as I would like.

I have been following this thread.

Greg Z wrote:

"- All of this is just a thought experiment if manufacturers aren't on board, 
willing to do the extra work, and see these extra costs as providing value. We 
shouldn't use any AllJoyn resources until they weigh in."

Honeywell would benefit from having an AllSeen/AllJoyn ROT.

Being able to verify the manufacturer, make, model, and serial number of a 
device has value in many of the environments we operate in.

[DT] Can you given an example?  (The examples below don’t appear to me to 
require verifying manufacturer, etc.)

Here are two items of concern which cut across multiple use cases. These 
highlight why Honeywell needs a RoT and the associated support.

·         Increased liability for failure. Many IoT devices have a relatively 
low impact if they fail. (e.g. light bulbs and TVs). However, Honeywell 
produces products which could lead to serious consequences if they fail (CO 
detectors, smoke detectors and home security systems). Our potential liability 
is high and thus demands that we take steps to minimize the risk of failure due 
to poor security.

[DT] Agree, but if I understand the above scenario it just requires AlLJoyn 
security, it does not require an AlLSeen ROT.   In normal AllJoyn security, the 
owner of the device or network is the ROT.   Can you elaborate on what 
additional value an AllSeen ROT would have over-and-above the normal AllJoyn 
security model where the owner is the ROT?   Just trying to understand not 
pushing back.

[TRM > ]  The Rot enables automation and validation of the onboarding process. 
If the onboarding process is manual then there is an increased risk of human 
error. We must consider the case of an untrained homeowner who may not have 
read the user manual for a device before joining it to their network. Let’s dig 
into a (long) concrete example.

A homeowner buys a Vendor_X home security panel which includes intrusion 
detection (door sensors, motion detectors, etc.) and fire alarm. 
http://www.security.honeywell.com/hsc/products/intruder-detection-systems/control-panel/fire-burglary/<https://na01.safelinks.protection.outlook.com/?url=http%3a%2f%2fwww.security.honeywell.com%2fhsc%2fproducts%2fintruder-detection-systems%2fcontrol-panel%2ffire-burglary%2f&data=01%7c01%7cgregz%40microsoft.com%7c327b52da6bf8432db56308d2df2d619f%7c72f988bf86f141af91ab2d7cd011db47%7c1&sdata=BPmijDMxbNDUkejdAGMCDwTeSPqpk%2fqYfLBM0DQFPoA%3d<http://www.security.honeywell.com/hsc/products/intruder-detection-systems/control-panel/fire-burglary/%3chttps:/na01.safelinks.protection.outlook.com/?url=http%3a%2f%2fwww.security.honeywell.com%2fhsc%2fproducts%2fintruder-detection-systems%2fcontrol-panel%2ffire-burglary%2f&data=01%7c01%7cgregz%40microsoft.com%7c327b52da6bf8432db56308d2df2d619f%7c72f988bf86f141af91ab2d7cd011db47%7c1&sdata=BPmijDMxbNDUkejdAGMCDwTeSPqpk%2fqYfLBM0DQFPoA%3d>>
  These panels receive a sensor input from a local fire or smoke sensor and 
send the alarm to an alarm monitoring center (A.K.A. central station). The 
alarm monitoring center then contacts the local fire department. The DIY 
homeowner buys a smoke detector off the internet which claims to work with 
Vendor_X fire-burglary panels. In fact the device is not certified to work with 
the Vendor_X panel but the DIY homeowner does not know this. Unfortunately, 
there are counterfeit products sold over the internet. The fire-burglary panel 
is able to validate smoke detectors from Vendor_X because both the panel and 
smoke detector have a common Vendor_X root. However, if there is no common ROT, 
then the panel does not have an automated, authenticated means of determining 
if the smoke detector is what it claims to be. The homeowner, using a manual 
process completes the installation of the smoke detector.

Sometime later there is a fire in the home. The system (smoke detector & panel) 
fails to detect the fire and contact the monitoring center. The homeowner’s 
child dies in the fire. The home owner sues Vendor_X. Did Vendor_X take 
“reasonable care” to ensure that the smoke detector would work with the panel?

If there is a common RoT then the panel has the ability to check the device 
certificate in the smoke detector and accept or reject the smoke detector. The 
chance of human error is eliminated when the panel can use a common RoT to 
automatically validate the smoke detector. If public key technology is 
available in 2015 to address this issue and Vendor_X does not use a common RoT, 
what is the Vendor_X liability? What will the jury decide?

·         Scalable secure commissioning: Honeywell (and other vendors) need to 
consider the labor and risk of human error in large deployments. For example, 
consider the commissioning of a new apartment building having 100 units, with 
each unit having 4 smoke detectors. All of these are networked to a central 
building fire panel. What method of commissioning (claiming or On-boarding) 
these 400 smoke detectors would allow the installer to reliably and cost 
effectively register these devices with the building’s central fire panel? One 
could imagine taking pictures of bar codes on 400 nearly identical smoke 
detector boxes and manually maintaining the associations. Alternatively, one 
could imagine an automated PKI assisted mobile device application which 
incorporates claiming and commissioning the devices in a manner which is 
approved by the fire marshal. Many buildings (and homes) will contain devices 
from multiple vendors. Having an AllSeen RoT enables (but does not guarantee) a 
secure and scalable commissioning process.

[DT] Again, I didn’t follow what additional value an AllSeen ROT would have 
over-and-above the normal AllJoyn security model where the owner is the ROT.  
Can you elaborate?

[TRM > ]  Once the device has been brought into the owner’s network the owner’s 
RoT and associated security mechanisms can provide a lot of security 
protection. The threat the AllSeen wide RoT addresses is protecting the owner 
from joining an “incorrect” device to the network in a multi-vendor 
environment. The device may be incorrect because it is the wrong model or 
because it is a counterfeit. The AllSeen RoT provides value to the owner by 
giving the owner the ability to automatically and with confidence perform 
onboarding.

I encourage the group to think not just about the use cases and devices of 
today but imagine the world 5-10 years from now. Consider the issues we are 
facing today in migrating from IPv4 to IPv6 because we did not anticipate the 
scale of the network. If we do not provide the option today for vendors to 
participate in an AllSeen/AllJoyn wide PKI, then we may find that we hit a wall 
in terms of scalability, trust and automation.

What if AllSeen has an optional RoT? Vendors who value automation, reduced 
human error and management of risk can choose to participate by having their 
vendor CA signed by the RoT. Vendors selling low risk products (e.g. light 
bulbs) may choose to sell devices not traceable to the RoT. The person or 
process doing the onboarding then has the option of deciding the level of risk 
appropriate for their environment.

Bottom line - Honeywell does see real value in a having a globally verifiable 
"certificate".

Tom Markham
From: Brian Witten [mailto:[email protected]]
Sent: Tuesday, February 09, 2016 5:26 PM
To: Swinson, Ken <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; 
[email protected]; [email protected]; Markham, 
Tom <[email protected]>; [email protected]; Damon Kachur 
<[email protected]>
Subject: Re: [Allseen-core] Security 2.0 status, and proposals for improvements

Hi Ken,

Sadly I've been too distracted with the day job.  I know that Honeywell and 
CableLabs and Symantec are all aligned on the merits and need for Device 
Certificates managed by a responsible CA.  However, I'm stepping into 
international travel, and won't really be back until Feb 22.  Perhaps a good 
next step is for me to set up a call for that last week of February?  
Alternatively, if many of you will be at the RSA Conference, I could book a 
room for many of us to meet f2f there, and dial in the others?  Thank You 
Either Way -

Best,
Brian


On Feb 09, 2016, at 11:56 AM, "Swinson, Ken" 
<[email protected]<mailto:[email protected]>> wrote:
I do not have a business case to drive the mfg/device certificate use but I am 
willing to assist with documenting the use cases if someone steps forward to 
drive it.  I’m concerned that those who expressed interest are assuming that 
functionality is in place when it isn’t.

I’m looking forward to your proposal for the manifest changes.

From: Daniel Mihai [mailto:[email protected]]
Sent: Monday, February 08, 2016 6:28 PM
To: Swinson, Ken; Lioy, Marcello; 
[email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Cc: [email protected]<mailto:[email protected]>; Dominique Chanet 
(QEO LLC); Josh Spain; 
[email protected]<mailto:[email protected]>
Subject: RE: [Allseen-core] Security 2.0 status, and proposals for improvements

Thanks for the feedback Ken!

Some thoughts below:

From: Swinson, Ken [mailto:[email protected]]
Sent: Monday, February 8, 2016 3:04 PM
To: Daniel Mihai 
<[email protected]<mailto:[email protected]>>; Lioy, Marcello 
<[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Cc: [email protected]<mailto:[email protected]>; Dominique Chanet 
(QEO LLC) <[email protected]<mailto:[email protected]>>; Josh 
Spain <[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>
Subject: RE: [Allseen-core] Security 2.0 status, and proposals for improvements

Dan, Thanks for the list,

Has device certificates dropped from the active discussions?  I suspect code 
changes would be desirable regardless of the decision to move forward with a 
single root CA or multiple root CAs selected by individual OEMs.
[DM] This one is currently not on my radar at all. Please bring it up in the 
Core WG meetings if it’s important.


See inline for other comments.

Ken


From: Daniel Mihai [mailto:[email protected]]
Sent: Friday, February 05, 2016 2:11 PM
To: Lioy, Marcello; [email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Cc: [email protected]<mailto:[email protected]>; Dominique Chanet 
(QEO LLC); Swinson, Ken; Josh Spain; 
[email protected]<mailto:[email protected]>
Subject: RE: [Allseen-core] Security 2.0 status, and proposals for improvements


This is the current high-level wish list of improvements that we are 
investigating at MSFT. It is still unclear how many and when we’ll be able to 
implement them.



1. Add new Manifest format, more granular, as I proposed in the attached email.

                - Also consider removing the support for the problematic 15.09 
Manifest format

[kens]  I need clarification on what you are proposing.   Is it as Phil 
summarized in the e-mail you attached?   In which case then more granular 
refers to segmenting the manifest?  The 15.09 manifest should not be supported 
if it is replaced.

[DM] We will share our detailed proposal as soon as possible. We are currently 
reviewing one such proposal in our team – stay tuned.





2. Add XML-based public methods for Policy, Manifest, Manifest Template.

- Consider removing the current public methods, if they are superseded by the 
new methods

[kens] I agree with removing superseded methods that would fall under the 
developer preview for sec 2



3. Add StartManagement + EndManagement methods, called by a Security Manager to 
notify the target app that its Security settings are about to change.

[kens]  I’ll reply to your later email on the topic.



4. Add public C APIs corresponding to the most useful public C++ methods.



5. Significant changes to the implementation of the Key Store – to address 
ASACORE-2661, ASACORE-2664, and other shortcomings.



6. Make it easier to share a single Key Store between two processes running on 
a single machine.

[kens]  This should include making it easy for a security manager to determine 
that the key store is shared so that it can construct policies and manifests.

[DM] Agreed that would be a worthy goal. Let’s see how close we can get to that 
goal in 16.04. We’ll provide our detailed proposal as soon as we can.





7. Add a new SPEKE auth mechanism – similar but better than the current PSK.



8. Introduce the concept of “Recommended Security Level”. For example:

- Manufacturer of an  “org.Contoso.GunSafe” device might want to recommend that 
interface as accessible just to “Privileged” users

- Manufacturer of an  “org.Contoso.WirelessSpeaker” device might want to 
recommend that interface as “NonPrivileged”

- Manufacturer of an  “org.Contoso.LightSwitch” device might want to recommend 
that interface as “Unauthenticated”

[kens]  I like it.  Consider mapping the recommendation to the peer types for 
consistency.

[DM] Peer types would definitely be used by the Security Manager behind the 
scenes, to implement the Security Levels mentioned above:



“Unauthenticated” maps well to PEER_ALL.



I look at a particular implementation of Security Manager as having more 
flexibility when it wants to implement “NonPrivileged” and/or “Privileged” 
Levels. For example, we’re currently still debating in our team if we want to 
use just PEER_FROM_CERTIFICATE_AUTHORITY, or just PEER_WITH_MEMBERSHIP, or a 
combination of these two, for our particular scenarios. Each of these choices 
seems to have advantages and disadvantages.



9. Investigate the numerous active JIRA tickets related to Security and fix 
those that the Core WG decides are important.



If anyone has questions or other kind of feedback, please speak up.



Dan



-----Original Message-----
From: Lioy, Marcello [mailto:[email protected]]
Sent: Thursday, February 4, 2016 11:39 AM
To: [email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Cc: Daniel Mihai <[email protected]<mailto:[email protected]>>
Subject: RE: [Allseen-core] Security 2.0 status



There are some changes being proposed, however, I think it is developer API 
changes rather than the interface definitions on the wire.  I have CC'ed Dan 
Mihai who is driving the updates to Security 2.0.  He can correct or elaborate 
as appropriate.



-----Original Message-----

From: 
[email protected]<mailto:[email protected]>
 [mailto:[email protected]] On Behalf Of 
[email protected]<mailto:[email protected]>

Sent: Monday, January 25, 2016 2:08 AM

To: 
[email protected]<mailto:[email protected]>

Subject: [Allseen-core] Security 2.0 status





Hi everyone,



Is it possible to release the latest 15.09(a)-related Security 2.0 HLD ? As 
both the 
https://na01.safelinks.protection.outlook.com/?url=allseenalliance.org&data=01%7c01%7cDaniel.Mihai%40microsoft.com%7c1761def47b9b40ff54da08d32d9adcb0%7c72f988bf86f141af91ab2d7cd011db47%7c1&sdata=8t3KSB3ksxrk%2fFOPp7gcPWW%2btMdZsmovppbK0g3%2fwzQ%3d
 documentation and wiki seem to point to out-of-date documents.



Also, is the production-grade 16.04 Security 2.0 roadmap decided upon (e.g. 
things like interface description servers and like) ?



Thanks,

Ondrej T



_______________________________________________

Allseen-core mailing list

[email protected]<mailto:[email protected]>

https://na01.safelinks.protection.outlook.com/?url=https%3a%2f%2flists.allseenalliance.org%2fmailman%2flistinfo%2fallseen-core&data=01%7c01%7cDaniel.Mihai%40microsoft.com%7c1761def47b9b40ff54da08d32d9adcb0%7c72f988bf86f141af91ab2d7cd011db47%7c1&sdata=%2fZzxaUmnFC9opwKhE%2bnozXKB04TnGNL3Kxw%2bn3FtYtM%3d
_______________________________________________
Allseen-core mailing list
[email protected]
https://lists.allseenalliance.org/mailman/listinfo/allseen-core

Reply via email to