Last Call Review of draft-ietf-perc-srtp-ekt-diet-09
I am the assigned Gen-ART reviewer for this draft. The General Area
Review Team (Gen-ART) reviews all IETF documents being processed
by the IESG for the IETF Chair. Please treat these comments just
like any other last call comments.
For more information, please see the FAQ at
Reviewer: Christer Holmberg
Review Date: 2018-10-20
IETF LC End Date: 2018-11-01
IESG Telechat date: Not scheduled for a telechat
Summary: The document is well written, and easy to read. But, I think some things still need to be clarified. I also have some editorial modification suggestions.
The text in section 1 says that EKT removes the need for co-ordinating SSRCs, and that an endpoint can choose whatever value it wants.
However, I can't find any explanation on how EKT removes that need.
The text in section 1 says that EKT is not intended to replace key establishment mechanisms, but to "be used in conjunction".
However, there is no description on how that works in conjunction with existing mechanisms (e.g., together with SDP Offer/Answer),
and whether existing mechanisms need to be modified in order to work in conjunction with EKT.
Section 5.3 does say that the SDP O/A exchange is not affected. If that is true, then you need to describe how SSRC values etc signalled in SDP relate to values signalled using EKT, what happens if there are mismatches etc.
The text in section 1 says:
"However, there are several cases in which conventional signaling systems cannot easily provide all of the
I think some examples would be useful.
The text in section 1 says that providing SSRCs etc using signaling systems cause layer violations.
Is this layer violation described somewhere? If so, I think a reference would be useful.
The text in section 4.2 says:
"All of the received EKT parameter sets SHOULD be stored by all of the
participants in an SRTP session, for use in processing inbound SRTP
and SRTCP traffic."
What is the reason for SHOULD, instead of MUST? What happens if an endpoint does NOT store the EKT parameter sets?
The text in section 4.2.1 says:
"Outbound packets SHOULD continue to use the old SRTP Master Key for
250 ms after sending any new key."
What is the reason for SHOULD, instead of MUST?
The text in section 4.3 says:
"Section 4.2.1 recommends that SRTP senders continue using an old key
for some time after sending a new key in an EKT tag."
The text in section 4.2.1 contains a SHOULD, so it is more than a recommendation.
I think Section 1 should also indicate that EKT is an DTLS extension, similar to the Abstract. Otherwise, when you say that EKT is a way to distribute information, it is unclear why EKT doesn't violate layers.
I think that Section 2 (Overview) could be combined with Section 1 (Introduction). Introduction sections in RFCs typically provide both background, problem statement and an overview of the mechanism - and sometimes also the document structure.
The text in section 4.1 says:
"EKT MUST NOT be used in conjunction with SRTP's MKI (Master Key
Identifier) or with SRTP's <From, To> [RFC3711], as those SRTP
features duplicate some of the functions of EKT. Senders MUST NOT
include MKI when using EKT. Receivers SHOULD simply ignore any MKI
field received if EKT is in use."
I suggest to put this text in a dedicated Applicability section.
The text in section 4.1 says:
"Together, these data elements are called an EKT parameter set. Each
distinct EKT parameter set that is used MUST be associated with a
distinct SPI value to avoid ambiguity."
I suggest to defined "EKT parameter set" in the same way as the other terminology. I.e.,
"EKT parameter set: The parameters indicated by the SPI"
...or something like that.
For the section names of sections 4.2.1 and 4.2.2, I suggest to talk about sending/transmitting and receiving instead of inbound and outbound.
In section 4.3, I think the name of the section ("Implementation Notes") is a little confusing. Also, is there a reason why the text is not in section 4.2.1 and/or 4.2.2?
Sections 4.4 and 4.4.1 have the same section names. Please change one of them (e.g., change 4.4.1 to "Default Cipher").
I think the text in section 4.6 should be placed in the Applicability section I suggested earlier (Q4_1).
In section 5, I suggest to change the section name to "DTLS-SRTP Considerations".