Hello Chairs, Authors, and WG,

I have the following comments on draft-mglt-nvo3-geneve-security-requirements 
that needs to satisfied before the WG adoption:


1.       Section 2: Not all data centers environments have all the 
risks/threats highlighted in this document. The text in this section could be 
interpreted that all of the requirements in this document needs to be met, 
whereas in practice only some of the requirements (not all of these 
requirements) may  be applicable to a certain deployment(s). So some of the 
requirements would not be applicable in such scenarios. So rephrase the 
paragraphs in section 2 (and in section 4) that captures the intent that not 
all requirements are applicable in to all deployments, you could state 
something along the lines of the following: In certain data center environments 
operated by a cloud provider or a private enterprise or data centers where 
certain risks/threats highlighted in this document may not be applicable. Hence 
one or more of the requirements identified in this document may not be 
applicable to those use cases and the data center operator may choose to deploy 
solutions that are only relevant for their application(s).



2.       GEN-REQ1.  As a policy, a tenant may decide to encrypt the packets 
irrespective of the privacy mechanisms offered by the service provider. So in 
this scenario the Tenant system administrator will enforce the policy for data 
privacy in the tenant systems/network, if needed. So it is not the NVE 
responsibility in all cases to enforce privacy. Only if the overlay provider is 
offering data privacy as a service then the NVE would provide that as a 
service. So GEN-REQ1 NVE requirement should be changed to “MAY”.  This is 
consistent with the “MAY” in confidentiality REQ 11 (and REQ16) in the 
draft-ietf-nvo3-security-requirements document. In fact this is not unique to 
Geneve encap and could apply to even other encapsulations. So consider keeping 
this consistent with nvo3 security requirements document.


3.       GEN-REQ2: This could simply be stated as, “If best protection from 
traffic analysis is needed, the NVE MAY encrypt the entire Geneve payload that 
includes the inner headers and payload”.  Geneve supports multi-protocol 
payloads, so this is not just specific to IP, and applies to other protocols as 
well. The objective of this requirement is data privacy/confidentiality between 
NVEs for the entire Geneve payload carried over Geneve. Generally, the service 
provider may provide this assurance to tenants as part of their overlay 
service, if required by the tenant. This requirement is consistent with the 
confidentiality requirement “MAY” in draft-ietf-nvo3-security-requirements 
document and should cover all protocols including IP.

4.       Section 5.2.  This section introduces a Geneve Transit Node (GTN) that 
performs authentication on behalf of an NVE, and defines the requirements for 
this model. This is not part of the NVO3 architecture. Architecturally one 
could consider such GTN as part of the NVE.  In general, this section builds a 
case for authentication on partial headers or selected fields by prescribing a 
certain architecture and introducing new components/functions (in sections 5.2 
and 5.3).  Though this could be considered an optimization, however, it is not 
otherwise required to address a potential security threat(s). Hence 
requirements related this model and therefore partial authentication of partial 
header(s) for NVE or GTN would not be considered as a security requirement. 
This just adds undue requirements to NVE and GTNs, which otherwise is not 
needed to address any threat. GEN-REQ 4-7 and GEN-REQ 9 fall in this category 
and hence can be removed. The section mentions that these are consistent with 
draft-ietf-nvo3-security-requirements draft, however that document does not 
call for partial header(s)/field(s)/partial payload handling or a transit node 
performing authentication/encryption on behalf of an NVE.



* The requirement for authentication should be to ensure the packet originated 
from the right source NVE and it reached the right destination NVE. I think, 
this is the actual requirement that needs to be captured.



GEN-REQ8. It is not clear as to what threat is being addressed by requiring 
flow level granularity.  If communication between NVE to NVE need be 
encrypted/authenticated, then, at a minimum, security policy should be applied 
for the traffic between, for example, NVE A to NVE B or NVE A to NVE C, etc.  
Any granularity beyond that is not a requirement to address any threat.  Hence 
consider removing flow level granularity requirement in GEN-REQ8.


5.       Section 5.3: The main objective of requirements in this section is for 
protecting the traffic between the NVEs (privacy/confidentiality). Applying 
partial encryption because the tenant payload is already encrypted is not a 
security requirement for a threat, this can be considered more of an 
optimization.  A tenant requesting a service provider to offer data 
confidentiality as a service would expect the NVE to NVE links to be protected 
irrespective of whether the tenant chooses to use  encryption of its data at 
the application level or otherwise.   So adding requirements for partial 
headers/payload could not be considered as a security requirement to address a 
security threat, hence consider removing this requirement(s). GEN-REQ10-12 fall 
under this category.



* Hence in this section, the requirement for confidentiality/privacy between 
the NVEs need be captured. And in addition GEN-REQ13 is applicable that 
describes using different cryptographic keys to protect tunnels between NVEs.  
These two are consistent with REQ11 and REQ12 in 
draft-ietf-nvo3-security-requirements document.



6.       GEN-REQ14. Multicast-packets: This requirement is already captured in 
draft-ietf-nvo3-security-requirement document and also not unique to Geneve 
encapsulation. So this requirement could be removed as it is already covered. 
In most cases multicast is not supported in underlay networks and multiple 
unicast tunnels are used for supporting multicast at overlay level. Even if 
such a capability is desired for some use cases, this is addressed in REQ 13 of 
draft-ietf-nvo3-security-requirements document.



7.       Section 5.4: Anti-Replay: GEN-REQ15. Consider changing this 
requirement to “SHOULD” to be consistent with REQ10 in draft-ietf-nvo3 security 
requirements document. Also GTN is included in this requirement, and GTN may 
not always be responsible for this validation. So “SHOULD” is more appropriate 
for this requirement.



8.       GEN-REQ12: This requirement prescribes that a transit underlay 
intermediate node must be able to insert an encrypted Geneve Option in an 
encrypted/authenticated Geneve packet. Architecturally it makes it more 
appropriate to make such an intermediate node (that deals with encrypted Geneve 
packets) an NVE (even if it does not process the payload). This simplifies and 
keeps it consistent with the existing architecture and existing protocols could 
be used.

Thanks for your consideration.

Regards,
Ilango



From: nvo3 [mailto:[email protected]] On Behalf Of Bocci, Matthew (Nokia - 
GB)
Sent: Tuesday, February 27, 2018 1:04 AM
To: NVO3 <[email protected]>
Cc: [email protected]; [email protected]
Subject: [nvo3] Poll for WG adoption of 
draft-mglt-nvo3-geneve-security-requirements-03

This email begins a two-week poll for adoption of 
draft-mglt-nvo3-geneve-security-requirements-03 in the NVO3 working group.

Please review the draft and send any comments to the NVO3 list.
Please also indicate whether you support adoption of the draft as an NVO3 
working group document.

Simultaneously, we are also poling for any IPR that may apply to the draft.

Authors and contributors, are you aware of any IPR that applies to this draft?
If so, has this IPR been disclosed in compliance with IETF IPR rules (see RFCs 
3979, 4879, 3669 and 5378 for more details)?
If you are listed as a document author or contributor, please respond to this 
email stating of whether or not you are aware of any relevant IPR.
The response needs to be sent to the NVO3 WG mailing list. The document will 
not advance to the next stage until a response
has been received from each author and each contributor.

This poll closes on Wednesday 14th March 2018.

Regards

Matthew and Sam

_______________________________________________
nvo3 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nvo3

Reply via email to