Re: [lamps] [EXTERNAL] Re: [CMP Updates] Requesting a current CRL

[John Gray <[e-mail]>]

I think we should only need the DistributionPointName, as we need a way to give the CA the information it needs to be able to lookup the CRL from its internal list of issued CRL's.   I hope this should be enough information for a CA to distinguish which CRL is being requested.  Is this an assumption we should be able to make (that a CA would easily be able to determine a CRL by its DistributionPointName)?


   IssuingDistributionPoint ::= SEQUENCE {

        distributionPoint          [0] DistributionPointName OPTIONAL,

        onlyContainsUserCerts      [1] BOOLEAN DEFAULT FALSE,

        onlyContainsCACerts        [2] BOOLEAN DEFAULT FALSE,

        onlySomeReasons            [3] ReasonFlags OPTIONAL,

        indirectCRL                [4] BOOLEAN DEFAULT FALSE,

        onlyContainsAttributeCerts [5] BOOLEAN DEFAULT FALSE }


DistributionPointName ::= CHOICE {
        fullName                [0]     GeneralNames,
        nameRelativeToCRLIssuer [1]     RelativeDistinguishedName }


So is this good enough?

CRLStatus ::= SEQUENCE {
   crldpn    DistributionPointName,
   cRLNumber  INTEGER,
   thisUpdate    Time }


In David's (2):
CRLStatus ::= SEQUENCE {
   issuer            Name,
   cRLNumber  INTEGER,
   scope             IssuingDistributionPoint OPTIONAL }

Isn't the Issuer name redundant because we would be making the request to a specific CA (our CMP server), so that should not be needed as we are making a request to the issuer (or an RA acting on behalf of the issuer).   ThisUpdate is needed to indicate to the server to only send a CRL that is newer than the given thisUpdate time (otherwise send nothing).  This should gains the network efficiency we are looking to achieve.  The CRL number is to request specific CRL numbers when all other information is identical.

I like Russ's proposal of supporting multiple CRL's in a CRLStatusList as well (at the expense of a little more complexity).   Supporting CRLStatusList makes it more flexible in case there is a need to request multiple at once.  So I am fine with that flexibility.

CRLStatusList ::= SEQUENCE OF CRLStatus

CRLs ::= SEQUENCE OF CertificateList

   GenMsg:    {id-it TBD}, CRLStatusList
   GenRep:    {id-it TBD}, CRLs  |  < absent >


Cheers,

John Gray


From: Brockhaus, Hendrik <[e-mail]>
Sent: Tuesday, October 12, 2021 9:35 AM
To: David von Oheimb <[e-mail]>; Russ Housley <[e-mail]>; Lijun Liao <[e-mail]>
Cc: LAMPS WG <[e-mail]>; John Gray <[e-mail]>
Subject: AW: [lamps] [EXTERNAL] Re: [CMP Updates] Requesting a current CRL

Do we really need the whole IssuingDistributionPoint, or is the DistributionPointName sufficient?

Hendrik

Von: Spasm <[e-mail]<mailto:[e-mail]>> Im Auftrag von David von Oheimb
Gesendet: Dienstag, 12. Oktober 2021 10:12
An: Russ Housley <[e-mail]<mailto:[e-mail]>>; Lijun Liao <[e-mail]<mailto:[e-mail]>>
Cc: LAMPS WG <[e-mail]<mailto:[e-mail]>>; Brockhaus, Hendrik (T RDA CST SEA-DE) <[e-mail]<mailto:[e-mail]>>; John Gray <[e-mail]<mailto:[e-mail]>>
Betreff: Re: [lamps] [EXTERNAL] Re: [CMP Updates] Requesting a current CRL


Thanks for the clarifications.

So how about adapting the CRLStatus part of Russ' below proposal from

CRLStatus ::= SEQUENCE {
   crldpn    DistributionPointName,
   thisUpdate    Time }

to (1):

CRLStatus ::= SEQUENCE {
   scope             IssuingDistributionPoint OPTIONAL,
   thisUpdate    Time }

or to (2):

CRLStatus ::= SEQUENCE {
   issuer            Name,
   cRLNumber  INTEGER,
   scope             IssuingDistributionPoint OPTIONAL }

?
In this way the full scope information can be given if needed, while
for the typical (simple) cases, it should be sufficient to give either the thisUpdate value or the CRL issuer and number.

    David

On 11.10.21 23:42, David von Oheimb wrote:

So in the end the whole IssuingDistributionPoint

structure determines the scope.



  David

On 11.10.21 21:45, Russ Housley wrote:
Lijun:

Yes, I agree.  That us the reason for crldpn in the syntax that I suggested.

Russ

On Oct 11, 2021, at 3:26 PM, Lijun Liao <[e-mail]<mailto:[e-mail]>> wrote:

The URL specified in the extension CRL DP, should be able to identify the scope.


Russ Housley <[e-mail]<mailto:[e-mail]>> schrieb am Mo., 11. Okt. 2021, 20:11:
Hendrik:

As the introduction to Section 5 od RFC 5280 says, the scope can be based on "arbitrary local information".  Consider the example that I gave in my earlier response, where a CA uses multiple CRL distribution points.  One distribution point could be associated with NIST employees located in Boulder and a separate distribution point could be for NIST employees in Gaithersburg.  There are certainly other ways to do this, including indirect CRLs, but this is one straightforward approach.

Russ

On Oct 11, 2021, at 12:20 PM, Brockhaus, Hendrik <[e-mail]<mailto:[e-mail]>> wrote:

Russ

The introduction to Section 5 gives some examples on different scopes.
I du understand how to express theses scopes by using the flags David mentioned.

  *   "all certificates issued by CA X"
  *   "all CA certificates issued by CA X"
  *   "all certificates issued by CA X that have been revoked for reasons of key compromise and CA compromise",

But I have issues to understand how "all certificates issued to the NIST employees located in Boulder" would be expressed.
Can you give an example?

Hendrik

Von: Russ Housley <[e-mail]<mailto:[e-mail]>>
Gesendet: Montag, 11. Oktober 2021 17:26
An: David von Oheimb <[e-mail]<mailto:[e-mail]>>
Cc: Lijun Liao <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Brockhaus, Hendrik (T RDA CST SEA-DE) <[e-mail]<mailto:[e-mail]>>; John Gray <[e-mail]<mailto:[e-mail]>>
Betreff: Re: [lamps] [EXTERNAL] Re: [CMP Updates] Requesting a current CRL

David:

These flags are clearly part of the scope.  The use of CRL distribution points in certificates can be used to divide the population of certificates across many different CRLs.

Russ

On Oct 11, 2021, at 11:16 AM, David von Oheimb <[e-mail]<mailto:[e-mail]>> wrote:

Interesting discussion!
It looks to me that the currently open issue boils down to uniquely identifying the (latest) CRL information currently known to the client.
Even when scoped, indirect, and delta CRLs need to be taken into account,
according to https://datatracker.ietf.org/doc/html/rfc5280#section-5.2.3<https://urldefense.com/v3/__https:/eur01.safelinks.protection.outlook.com/?url=https*3A*2F*2Fdatatracker.ietf.org*2Fdoc*2Fhtml*2Frfc5280*23section-5.2.3&data=04*7C01*7Chendrik.brockhaus*40siemens.com*7C4ad01b5d9113414a477f08d98d580223*7C38ae3bcd95794fd4addab42e1495d55a*7C1*7C0*7C637696232756680209*7CUnknown*7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0*3D*7C3000&sdata=KqGxNtGOOfOJukKWAIWwBhiljMUJMIwKvLEd9zzefcA*3D&reserved=0__;JSUlJSUlJSUlJSUlJSUlJSUlJSU!!FJ-Y8qCqXTj2!I9hCPwvpYvYsKczRCM_y70-117Xe4Yvw5aB2pzCAZ768NKnbCwz8aaL09vCbUq19$>
apparently the following data should be sufficient for unique identification:
    CRL issuer, scope, and CRL Number,
or am I missing something?
And how is the scope defined/encoded?
I'm having a hard time getting this nailed down from RFC 5280 and other sources,
but according to https://datatracker.ietf.org/doc/html/rfc5280#section-5.2.5
<https://urldefense.com/v3/__https:/eur01.safelinks.protection.outlook.com/?url=https*3A*2F*2Fdatatracker.ietf.org*2Fdoc*2Fhtml*2Frfc5280*23section-5.2.5&data=04*7C01*7Chendrik.brockhaus*40siemens.com*7C4ad01b5d9113414a477f08d98d580223*7C38ae3bcd95794fd4addab42e1495d55a*7C1*7C0*7C637696232756690198*7CUnknown*7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0*3D*7C3000&sdata=Xi1FhU*2BLZ45xBsdwSQPlUPvwF3zIrv0*2F*2FMHy4nr7sm0*3D&reserved=0__;JSUlJSUlJSUlJSUlJSUlJSUlJSUlJSU!!FJ-Y8qCqXTj2!I9hCPwvpYvYsKczRCM_y70-117Xe4Yvw5aB2pzCAZ768NKnbCwz8aaL09sF5N2-s$>it looks like the scope is the combination of the following fields:

        onlyContainsUserCerts      [1] BOOLEAN DEFAULT FALSE,

        onlyContainsCACerts        [2] BOOLEAN DEFAULT FALSE,

        onlySomeReasons            [3] ReasonFlags OPTIONAL,

        indirectCRL                [4] BOOLEAN DEFAULT FALSE,

        onlyContainsAttributeCerts [5] BOOLEAN DEFAULT FALSE }
right?
    David

On 11.10.21 08:28, Brockhaus, Hendrik wrote:
Russ

Thank you for this proposal. It looks straight forward.
I will need to dig a little deeper into the partitioning of CRLs and check the requirements of our use case for CRL retrieval via CMP to better understand the complexity needed.

Hendrik

Von: Russ Housley <[e-mail]><mailto:[e-mail]>
Gesendet: Samstag, 9. Oktober 2021 21:07
An: Lijun Liao <[e-mail]><mailto:[e-mail]>
Cc: Brockhaus, Hendrik (T RDA CST SEA-DE) <[e-mail]><mailto:[e-mail]>; [e-mail]<mailto:[e-mail]>; von Oheimb, David (T RDA CST SEA-DE) <[e-mail]><mailto:[e-mail]>; John Gray <[e-mail]><mailto:[e-mail]>
Betreff: Re: [lamps] [EXTERNAL] Re: [CMP Updates] Requesting a current CRL

This is an interesting observation, but it does not seem to cover a certificate with multiple CRL distribution points, indirect CRLs, and delta CRLs.  I'm not sure we want all of that complexity here.  That said, it does make sense to me to list a distribution point name and the thisUpdate for each one.

CRLStatusList ::= SEQUENCE OF CRLStatus

CRLStatus ::= SEQUENCE {
   crldpn    DistributionPointName,
   thisUpdate    Time }

CRLs ::= SEQUENCE OF CertificateList

   GenMsg:    {id-it TBD}, CRLStatusList
   GenRep:    {id-it TBD}, CRLs  |  < absent >

Russ


Any email and files/attachments transmitted with it are confidential and are intended solely for the use of the individual or entity to whom they are addressed. If this message has been sent to you in error, you must not copy, distribute or disclose of the information it contains. Please notify Entrust immediately and delete the message from your system.