Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [2600] Updated FoPP-A version 29a



Hi Brian -

 

I have several comments on the latest P2600.1.D29 PP draft now that I actually have time to review the document. Note that many of the comments cover multiple PPs but I’ll only bring it up for the first instance I found it; you know what is repetitive better than I do anyway.

 

I know you probably will not be able to address these before next week's meeting; I'm hoping we can discuss them at the meeting or off-line as necessary:

1.    User Entity Definitions – The wording for the U.ORIGINATOR entity doesn’t seem to be correct. Do you mean to say “A User (you forgot to capitalize ‘user’) who is permitted to perform Normal User Document processing function…”. By the way, I believe we deleted the concept of a ‘normal user’ last meeting; if you are going to use that term here doesn’t that mean you now have to define what you mean by a ‘normal user’?

2.    User Entity Definitions – Remove the word ‘Administrator’ from the definition of the TU.ADMINISTRATOR entity.

3.    Object (Assets) Entity Definitions, line 15 – You seem to inconsistently refer to ‘states’ and ‘object-states’ here. Are these meant to be the same thing (if so use consistent terminology) or are they two different things (if so you’ll need to define what they both mean)?

4.    Object (Assets) Entity Definitions Table – I’m having a hard time understanding what you mean by ‘pending’ in this definition. I think you are trying to convey the concept that the job is being stored in the TOE either waiting to be processed or is being processed; if it’s being processed there is nothing ‘pending’ about the job. If this interpretation is correct, then I think a better definition might “Contained in the TOE either waiting for (awaiting?) or during performance of a job”. What do you think?

5.    Along the same lines, I don’t understand in all of the entity definitions why you use the term “contained in the TOE”. Documents and data are stored either temporarily or permanently in the TOE but that doesn’t mean that the TOE contains these documents or data (I am assuming you mean “contain” in the context of “hold” rather that “incorporate”). You could argue contain fits when you are talking about persistent storage, but it certainly doesn’t make sense when you are talking about temporary storage. I think you need to come up with a better term here that conveys both the temporary and permanent nature of the data being stored (I don’t have any good candidates at the moment but maybe I can give it some thought before next Tuesday).

6.    I have the same type of comment like #5 for data items being “present” in the TOE. I’m not sure what that really means. Seems to me in comparing D.PROT.atRest with D.PROT.atRest except for the data being protected the two data items are the same. In that case why wouldn’t they be in the same state – why create two different states here?

7.    P2600.1-PRT PP, page 14 – I think in your rationale for not including D.DOC.onServer you meant to say “This TOE does not store documents…”

8.    P2600.1-PRT PP, page 15 – Minor nit – to be consistent with the main P2600 body, in the rational for D.DOC.isDeleted you should use the term ‘nonvolatile’ instead of ‘non-volatile’.

9.    P2600.1-PRT PP, page 15 – The example for the TSF Asset D.PROT.atRest indicates that we are protecting this TSF data from “unauthorized alteration”. This seems to contradict the definition of D.PROT.atRest on page 10 which states that this data must be protected from “unauthorized modification”.

10.P2600.1-PRT PP, page 16 – The description of the P.SOFTWARE.VERIFICATION OSP doesn’t make sense to me. What you should be verifying here in my view is that the currently installed software in the TOE, and not the currently installed TOE, has remained consistent with the delivered and (initially) installed software. Suggest changing the wording to say “…verify that the currently installed software in the TOE has remained consistent with the delivered and initially installed software.”

11. P2600.1-PRT PP, page 20 – In the PRT Audit data requirements table, it’s not clear to me what an audit storage failure is in the context of FAU_STG.4. As I read this SFR, what we probably should be reporting each time the audit log becomes full – this is not an audit log failure which I would associate with something like not recording an event it was supposed to record or not writing the audit log file properly in a readable format when requested to do so.

12. P2600.1-PRT PP, page 20 – Some additional suggested minimum audit log entries:

·         FAU_SAR.1 – Also record when the audit log is downloaded and by whom (user_id)

·         FMT_SMR.1 – record administrator log-in attempts

·         FMT_SMF.1 – record whenever someone changes (or attempts to change) the SA password/PIN

·         FDP_ACF.1 - you should also record the job type so the job initiation and completion information makes sense and can be properly analyzed

·         FTA.SSL.3 - you should also record when a TSF-initiated session is first initiated

·         FIA_AFL.1 - you should also record the user id of anyone who reaches the failure limit (took me a couple of times before the wording of this element made sense) plus the number of log-in attempts by user id.

13. P2600.1-PRT PP, page 28 – I think it makes more sense that the definition of the PRT Management function for FAU_UID.1 should be management of “user actions” instead of “user identities”. I also am not sure I understand for FAU_SAR.1 who an “authorized audit user” is because this is the first time you use that term. Do you mean an “authorized audit record User” here (paraphrasing what is in FAU_SAR.1)?

14. For the P2600.1-DSR PP, can you provide a rationale why the software verification requirements and associated objectives/policy needs to be included? I understand why it was added to P2600.-PRT, -SCN, -CPY & -FAX but in this case we are talking about document storage and retrieval – why would we worry that the TOE software is verified as being correct when our concern here is around ensuring proper access and control of documents stored in the TOE – I just don’t see the connection. By the way, in the definition for the DSR PP, given the distinction you made in the previous PPs about data stored before, during or after processing of a job I think you need to clarify for DSR what stored documents the PP is referring to. From previous conversations my understanding was that the DSR PP only applies to documents stored after processing of a job; whether or not that is true the description of the DSR PP has to clarify what documents it applies to.

15. P2600.1-NVS PP, page 115 – Under the NVS TOE Model I’m curious why this PP only applies to User Document Data that is stored in NVS; that may be of primary concern here but we also have to be concerned that Confidential and Protected data like secrets that might be stored in NVM isn’t retrieved either.

16. P2600.1-NVS PP, page 118 – The NVS OSP section in the application note refers to the P.ADMIN and P.AUDIT policies; you probably should use the full names for the two referenced OSPs; otherwise it would imply there are other missing Admin and Audit OSPs. Same comment holds for the NVS Assumptions and NVS Security Objectives.

17. P2600.1-NVS PP, page 122 – I was wondering why you didn’t include any minimum audit data requirements associated with FDP_RIP.1 and the cryptography SFRs – these are the real SFR differences between NVS and the other PPs. As a minimum I would think you would want to include in the audit log some event associated with deletion of the applicable data in NVS and some event associated with the use of encryption (like “use of crypto libraries”). I’ll give some thought as to how to word these for next week’s meeting.

18. P2600.1-NVS PP, page 130 – Similar to the previous comment and for the same basic rationale, I was wondering why you didn’t include a management function recommendation associated with FDP_RIP.1 like ‘Management of User Document Data deletions’ and with the cryptographic SFRs such as “.

19. P2600.1-NVS PP, page 134 – The table on page 133 indicates that for O.DOC.atRest.NO_DIS SFRs FIA_UID.1, FMT_MSA.2, FMT_MTD.1 & FMT_SMR.1 help satisfy this objective. However, in the associated table on page 134 only the FMT_MTD.1 & FMT_SMR.1 are included in the discussion of O.DOC.atRest.NO_DIS. Same comment applies to O.PROT.atRest.NO_ALT & O.CONF.atRest.NO_ALT.

20. P2600.1-SMI PP, page 137 – There seems to be an type here in Item b) under ‘Major Security Features’ – shouldn’t that item read “The SMI TOE can protect Assets from disclosure or alteration when such Assets are exchanged…”

21. P2600.1-SMI PP, page 141 – In the table on this page you forgot to include the check for P.SMI.MEDIATE to satisfy O.SMI.MEDIATE. By the way, is it permissible that an OSP and associated security objective be exactly the same; to me the wording should be different to reflect the difference between a policy and an objective. Curious why they were worded the same.

22. P2600.1-SMI PP, page 154 – The table on page 154 indicates that for O.SMI.MEDIATE SFRs FDP_IFC.1 & FDP_IFF.1 help satisfy this objective. However, in the associated table on page 155 only the FDP_IFC.1 SFR is included in the discussion of O.SMI.MEDIATE.

23. P2600.1-SMI PP, page 154 – The table on page 154 indicates that for O.CONF.atRest.NO_DIS SFRs FIA_UID.1, FMT_MSA.1, FMT_MSA.3, FMT_MTD.1, FMT_SMF.1 & FMT_SMR.1 help satisfy this objective. However, in the associated table on page 134 only the FMT_MTD.1 & FMT_SMR.1 are included in the discussion of O.CONF.atRest.NO_DIS. Same comment applies to O.CONF.atRest.NO_ALT & O.PROT.atRest.NO_ALT.

24. P2600.1-SMI PP, page 154 – The O.USER_AUTHORIZED objective shown on the table on page 54 is not covered in the associated table on page 155.

I’ll check the Annex A Glossary later to see if anything is missing.

I hope these comments help. By the way, can you send me the word version of your latest draft so I can incorporate the applicable changes into the PP-E draft?

See you in Toronto.

 

Al Sukert

Product Security Specialist - XOG Product Security Office

XOG Export Control Coordinator

Xerox Certified Green Belt

Device Administration Program Manager

8*227-1413 or 585-427-1413

FAX: 585-427-6599

Alan.Sukert@xerox.com 

-----Original Message-----
From: Brian Smithson [mailto:Brian.Smithson@RICOH-USA.COM]
Sent: Tuesday, August 14, 2007 10:11 PM
To: STDS-2600@listserv.ieee.org
Subject: [2600] Updated FoPP-A version 29a

 

An updated FoPP-A has been posted here (with lots of change marks):

http://grouper.ieee.org/groups/2600/drafts/ProtectionProfiles/P2600.1-29a.

pdf

And here (without change marks):

http://grouper.ieee.org/groups/2600/drafts/ProtectionProfiles/P2600.1-29a_

ncb.pdf

 

You can also see the master asset/threat/objective/SFR worksheet here:

http://grouper.ieee.org/groups/2600/presentations/Toronto2007/P2600master-

29a.xls

 

 

 

There is much more blood on the pages than I had imagined, but nearly all

of it is the result of comments and action items from Bellevue's meeting.

Here is a summary of the changes:

 

AI#362 Added policy, objective, and SFRs for self test per Carmen's "case

2" proposal. I took the liberty of changing P.SOFTWARE.INTEGRITY to

P.SOFTWARE.VERIFICATION, and O.SOFTWARE.INTEGRITY to O.SOFTWARE.VERIFIED,

for consistency with the parts of speech in other P's and O's. The

policy/etc was added to all TOEs, even though it is somewhat redundant to

do so, because that will ensure that the verification function applies to

the software in each TOE. I think this will apply to environments A, B,

and C, but we didn't discuss that at the Bellevue meeting.

 

AI#368 Moved T.FUNC.STORED (job logs) out of User Function Data and into

TSF Protected Data. This seemed to work OK.

 

AI#369 Added an app note to the General model to say that not all assets

and asset-states defined in the General model apply to all TOEs. Added a

table in each TOE giving a rationale for which asset-states are not

considered in that TOE.

 

AI#371 Changed the wording of SFPs to say that a TOE _requires_

identification and authentication, not that it necessarily performs it.

This (I hope) leaves the door open to either perform I&A in the TOE or

outside of the TOE.

 

AI#372, #374, #375 Changed state names and state structure to reduce

confusion about what is temporary and what is stored. Refer to

http://grouper.ieee.org/groups/2600/email/msg00879.html for details.

 

AI#377 Added some clarification to the SMI TOE (and also the NVS TOE,

which had a similar issue) which says that their operations are being

performed on behalf of other Subjects. However, I think that the data is

inside of the TOE boundary.

 

AI#378 Removed P.COMMS.NO_BRIDGE and P.COMMS.NO_PROXY (and their

objectives) and replaced with P.SMI.MEDIATION and O.SMI.MEDIATED (same set

of SFRs).

 

AI#379 Removed FIA_UAU.6 from all TOEs. I could not think of a good

justification to keep it.

 

AI#382 Added back "detection" to FAU_STG.1.2 by restoring the original

unmodified SFR.

 

AI#383 Split DIS and ALT into different rows in all rationale tables.

 

AI#384 Created audit requirement and recommendation tables, and modified

FAU_GEN.*.

 

AI#385 Created management recommendation tables, and modified FMT_SMF.1.

 

AI#390 Audit event tables were created and should provide enough

clarification.

 

One item was NOT completed:

 

AI#380 I think that the need for SSL.3 does not need to be explicitly

stated in P*.AUTHORIZATION. Its supporting role is described in the

SFR-Objective rationale tables. The same thing also applies to FIA_AFL.1,

FIA_UAU.7, and others. Adding specific requirements would make those

policies (and their objectives) unwieldly.

 

 

Regards,

--

Brian Smithson

Project Manager

PMP, SSCP, CISSP, CISA, ISO 27000 PA

Advanced Imaging and Network Technologies

Ricoh Americas Corporation

(408)346-4435

 

New address:

10460 Bubb Road

Cupertino, CA 95014-4150