[Docs] [txt|pdf] [Tracker] [WG] [Email] [Diff1] [Diff2] [Nits]

Versions: (draft-stillman-rserpool-threats) 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 RFC 5355

Internet Engineering Task Force                       Maureen Stillman
INTERNET DRAFT                                        Ram Gopal
                                                      Senthil Sengodan
                                                      Nokia
                                                      Erik Guttman
                                                      Sun Microsystems
                                                      Matt Holdrege
                                                      Sonus Networks
                                                      12 December 2002
expires June 12, 2002

      Threats Introduced by Rserpool and Requirements for Security
                          in response to Threats
                 <draft-ietf-rserpool-threats-00.txt>



Status of This Memo


   This document is an Internet-Draft and is in full conformance with
   all provisions of Section 10 of RFC2026 [RFC2026].  Internet-Drafts
   are working documents of the Internet Engineering Task Force (IETF),
   its areas, and its working groups.  Note that other groups may also
   distribute working documents as Internet-Drafts.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at
   any time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at:
        http://www.ietf.org/ietf/1id-abstracts.txt
   The list of Internet-Draft Shadow Directories can be accessed at:
        http://www.ietf.org/shadow.html.


Abstract

   This Internet draft is an attempt to describe security threats
   against the Rserpool protocol.  This draft presents requirements for
   a security solution to thwart these threats in environments where it
   is likely to be deployed.  The threats and requirements identified
   herein and the document should be considered as work in progress.













Stillman, et al.                                                [Page 1]

 Internet Draft     Threats Introduced by Rserpool     12 December 2002




                                Contents


Status of This Memo                                                    1
Abstract                                                               1
1. Introduction                                                        3
     1.1 Definitions  . . . . . . . . . . . . . . . . . . . . . . .    3
2. Threats                                                             4
     2.1  PE Registration/Deregistration flooding . . . . . . . . .    4
     2.2  PE Registration/Deregistration flooding . . . . . . . . .    4
     2.3  PE Registration/Deregistration spoofing . . . . . . . . .    4
     2.4  PE Registration/Deregistration unauthorized . . . . . . .    5
     2.5  Malicious ENRP server joins the group of legitimate ENRP
                          servers . . . . . . . . . . . . . . . . .    5
     2.6  Registration/deregistration with malicious ENRP servers .    5
     2.7  Malicious ENRP Name Resolution .. . . . . . . . . . . . .    5
     2.8  Malicious node performs a replay attack.. . . . . . . . .    6
     2.9  Re-establishing PU-PE security during failover. . . . . .    6
     2.10 Integrity . . . . . . . . . . . . . . . . . . . . . . . .    6
     2.11 Data Confidentiality  . . . . . . . . . . . . . . . . . .    6
     2.12 ENRP Server Discovery . . . . . . . . . . . . . . . . . .    7
     2.13 Application security  . . . . . . . . . . . . . . . . . .    7
3.  Security Considerations . . . . . . . . . . . . . . . . . . . .    8
     3.1 Scenarios for using TLS with Rserpool  . . . . . . . . . .    8
       3.1.1  PE - ENRP security  . . . . . . . . . . . . . . . . .    9
         3.1.1.1 Scenario A - TLS only  . . . . . . . . . . . . . .   10
         3.1.1.2 Scenario B - TLS plus alternate method for client
                 authentication . . . . . . . . . . . . . . . . . .   10
         3.1.1.3 Open issues for ENRP-PE security . . . . . . . . . . 11
       3.1.2 End user-ENRP security . . . . . . . . . . . . . . . . . 11
     3.2 Scenarios for using IPsec with Rserpool  . . . . . . . . . . 11
       3.2.1 Scenario A - PU to ENRP server . . . . . . . . . . . . . 12
       3.2.2 Scenario B - PE to ENRP server . . . . . . . . . . . . . 13
4.  References  . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . . 14
6.  Author's addresses  . . . . . . . . . . . . . . . . . . . . . . . 15

















Stillman, et al.                                               [Page 2]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

1. Introduction

RSERPOOL provides a session layer for robustness and performance.  The
session layer function may redirect communication transparently to
upper layers.  This alters the direct one-to-one association between
communicating endpoints which typically exists between clients and
services.   In particular, secure operation of protocols often relies
on assumptions at different layers regarding the identity of the
communicating party and the continuity of the communication between
endpoints.  Further, the operation of RSERPOOL itself has security
implications and risks.  The session layer is organized and operates
dynamically which imposes additional concerns for the overall security
of the end-to-end application.  This document explores the security
implications of RSERPOOL, both due to its own functions and due to its
being interposed between applications and transport interfaces.

This draft is modeled after [MIPv6 threats] which is a threat analysis
document for Mobile IP V6.

1.1 Definitions

This document uses the following terms:

     ENRP Endpoint Name Resolution Protocol:
          Within the operational scope of Rserpool, ENRP defines the
          procedures and message formats of a distributed fault-tolerant
          registry service for storing, bookkeeping, retrieving, and
          distributing pool operation and membership information.

     ASAP Aggregate Server Access Protocol:
          A session layer protocol which uses the Endpoint Name
          Resolution Protocol (ENRP) to provide a high
          availability name space.  ASAP is responsible for the
          abstraction of the underlying transport technologies, load
          distribution management,fault management, as well as the
          presentation to the upper layer (i.e., the ASAP user) a
          unified primitive interface.

     Operation scope:
          The part of the network visible to pool users by a specific
          instance of the reliable server pooling protocols.

     Pool (or server pool):
          A collection of servers providing the same application
          functionality.

     Pool handle (or pool name):
          A logical pointer to a pool. Each server pool will be
          identifiable in the operation scope of the system by a unique
          pool handle or "name".

     ENRP namespace (or namespace):
          A cohesive structure of pool names and relations that may be
          queried by an internal or external agent.


Stillman, et al.                                                [Page 3]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

     Pool element (PE):
          A server entity that runs ASAP and has registered to a pool.

     Pool user (PU):
          A server pool user that runs ASAP. Note, a PU can also be a
          PE if it has registered itself to a pool.

     ENRP namespace server (or ENRP server):
          Entity which runs ENRP and is responsible for managing and
          maintaining the namespace within the operation scope.

2. Threats

2.1  PE Registration/Deregistration flooding

Threat:  A malicious node could send a stream of false
registrations/deregistrations on behalf of non-existent PEs to ENRP
servers at a very rapid rate and thereby create unnecessary state in an
ENRP server.

Effect: Corrupting the name server database and/or disabling the
Rserpool discovery and naming function.

Requirement:  An ENRP server that receives a registration/deregistration
should not create or update state information until it has authenticated
the PE.

2.2 PE Registration/Deregistration flooding

Threat: A malicious node or PE could send a stream of
registrations/deregistrations that are unauthorized to
register/deregister - to ENRP servers at a very rapid rate and thereby
create unnecessary state in an ENRP server.

Effect: Corrupting the name server database and/or disabling the
Rserpool discovery and naming function.

Requirement:  An ENRP server that receives a registration/deregistration
should not create or update state information until the authorization of
the registering/de-registering entity is verified.


2.3  PE Registration/Deregistration spoofing

Threat:  A malicious node could send false registrations/deregistrations
to ENRP servers concerning a legitimate PE thereby creating false state
information in the ENRP servers.

Effect: Misinformation in the ENRP server concerning a PE would get
propagated to other ENRP servers thereby corrupting the ENRP database.

Requirement:  An ENRP server that receives a registration/deregistration
should not create or update state information until it has authenticated
the PE.


Stillman, et al.                                                [Page 4]

Internet Draft     Threats Introduced by Rserpool      12 December 2002


2.4  PE Registration/Deregistration unauthorized

Threat:  A PE who is not authorized to join a pool could send
registrations/deregistrations to ENRP servers thereby creating false
state information in the ENRP servers.

Effect: Misinformation in the ENRP server concerning a PE would get
propagated to other ENRP servers thereby corrupting the ENRP database.

Requirement:  An ENRP server that receives a registration/deregistration
should not create or update state information until it has authorized
the requesting entity.

2.5 Malicious ENRP server joins the group of legitimate ENRP servers

Threat: Malicious ENRP server joins the group of legitimate ENRP servers
with the intent of propagating inaccurate updates to corrupt the ENRP
database.

Effect: Inconsistent ENRP database state.

Requirement: Mutual authentication of ENRP servers.

2.6 Registration/deregistration with malicious ENRP server

Threat: A PE unknowingly registers/deregisters with malicious ENRP
server.

Effect: Registration might not be properly processed or ignored.

Requirement: PE needs to authenticate the ENRP server.


2.7 Malicious ENRP Name Resolution

Threat: The ASAP protocol receives a name resolution response from an
ENRP server, but the ENRP server is malicious and returns random IP
addresses or an inaccurate list in response to the pool handle.

Effect: PU application communicates with the wrong PE or is unable to
locate the PE since the response is incorrect in saying that a PE with
that name did not exist.

Requirement: ASAP needs to authenticate the ENRP server.











Stillman, et al.                                                [Page 5]

Internet Draft     Threats Introduced by Rserpool      12 December 2002


2.8 Malicious node performs a replay attack

Threat: A malicious node could replay the entire message previously sent
by a legitimate entity. This could create false/unnecessary state in the
ENRP servers when the replay is for registration/de-registration or
update.

Effect: False/extra state is maintained by ENRP servers

Requirement:  Care should be taken to prevent replay attacks.

2.9 Re-establishing PU-PE security during failover

Threat: PU fails over from PE A to PE B.  In the case that the PU had a
trusted relationship with PE A, then the PU will likely not have the
same relationship established with PE B.

Effect: If there was a trust relationship involving security context
between PU and PE A, the equivalent trust relationship will not exist
between PU and PE B.  This will violate security policy.

Requirement: Either notify the application when fail over occurs so the
application can take appropriate action to establish a trusted
relationship with PE B OR reestablish the security context
transparently.

2.10  Integrity

Threats:
a) ENRP response to name resolution is corrupted during transmission
b) ENRP peer messages are corrupted during transmission
c) PE sends update for values and that information is corrupted during
transmission

Effect: ASAP receives corrupt information for pool handle resolution
which the PU believes to be accurate.

Requirement: Integrity mechanism needed.

2.11 Data Confidentiality

Threat: An eavesdropper capable of snooping on fields within messages in
transit, may be able to garner information such as topology/location/IP
addresses etc. that may not be desirable to divulge.

Effect: Information that an administrator does not wish to divulge are
divulged.

Requirement: Provision for Data confidentiality service.






Stillman, et al.                                                [Page 6]

Internet Draft     Threats Introduced by Rserpool      12 December 2002



2.12 ENRP Server Discovery

Threat A thwarting successful discovery: When a PE wishes to register
with an ENRP server, it needs to discover an ENRP server. An attacker
could thwart the successful discovery of ENRP server(s) thereby inducing
the PE to believe that no ENRP server is available. For instance, the
attacker could reduce the returned set of ENRP servers to null or a
small set of inactive ENRP servers.

Threat B: A similar thwarting scenario also applies when an ENRP server
or ASAP on behalf of a PU needs to discover ENRP servers.

Threat C: Spoofing successful discovery:  An attacker could spoof the
discovery by claiming to be a legitimate ENRP server.  When a PE wishes
to register, it finds the spoofed ENRP server.

Threat D: A similar spoofing scenario also applies when an ENRP server
or ASAP on behalf of a PU needs to discover ENRP servers.

Effect A: A PE that could have been in an application server pool does
not become part of a pool. The PE does not complete discovery operation.
This is a DOS attack.

Effect B: An ENRP server that could have been in an ENRP server pool
does not become part of a pool.  A PU is unable to utilize services of
ENRP servers.

Effect C,D: This malicious ENRP would either misrepresent, ignore
or otherwise hide or distort information about the PE to subvert
RSERPOOL operation.


Requirement: Discovery phase needs to be authenticated.


2.13  Security State for Applications

The security context of an application is a subset of the overall
context, and context or state sharing is explicitly out-of-scope for
RSerPool. Because RSerPool does introduce new security vulnerabilities
to existing applications application designers employing RSerPool should
be aware of problems inherent in failing over secured connections.
Security services necessarily retain some state and this state may have
to be moved or re-established. Examples of this state include
authentication or retained ciphertext for ciphers operating in cipher
block chaining (CBC) or cipher feedback (CFB) mode. These problems must
be addressed by the application or by future work on RSerPool.

Requirement: None at this time.  Future Rserpool work may address this
issue.




Stillman, et al.                                                [Page 7]

Internet Draft     Threats Introduced by Rserpool      12 December 2002


3. Security Considerations for Rserpool

Due to varying requirements and multiple use cases of Rserpool, we point
out two basic security protocols, IPsec and TLS. We specifically
do not discuss whether one security protocol would be preferred over the
other.  This choice will be made by designers and network
architects based on system requirements.

For networks that demand IPsec security, implementations MUST support
draft-ietf-ipsec-sctp-02.txt which describes IPsec-SCTP.  IPsec is
two layers below RSerPool. Therefore, if IPsec is used for securing
Rserpool, no changes or special considerations need to be made to
Rserpool to secure the protocol.

For networks that cannot or do not wish to use IPsec and prefer instead
TLS, implementations MUST support TLS with SCTP as described in
draft-ietf-tsvwg-tls-over-sctp-00.txt or TLS over TCP as described in
RFC 2246.  When using TLS/SCTP we must ensure that RSerPool does
not use any features of SCTP that are not available to an TLS/SCTP user.
This is not a difficult technical problem, but simply a
requirement. When describing an API of the RSerPool lower layer we have
also to take into account the differences between TLS and SCTP.
This is also not difficult, but it is in contrast to the IPsec solution
which is transparently layered below Rserpool.

Support for security is required for the ENRP server and the PEs.
Security support for the Rserpool end user is optional.  Note that
the end user implementation contains a piece of the Rserpool protocol --
namely ASAP -- whereby the pool handle is passed for name
resolution to the ENRP server and IP address(es) are returned.

The argument for optional end user security is as follows: If the user
doesn't require security protection for example, against
eavesdropping for the request for pool handle resolution and response,
then they are free to make that choice.  However, if
the end user does require security, they are guaranteed to get it due to
the requirement for security support for the ENRP server.
It is also possible for the ENRP server to reject an unsecured request
from the user due to its security policy in the case that it
requires enforcement of strong security.  But this will be determined by
the security requirements of the individual network design.


3.1 Scenarios for using TLS with Rserpool

This section describes security scenarios for two different parts of the
Rserpool protocol.  First, we examine the interaction between the PE
and ENRP server.  Next we examine a scenario for the end user (client
using ASAP) and ENRP server interaction.

Security provided by TLS includes authentication, confidentiality,
integrity, protection from replay attack and protection from downgrade
attack (that is coercing the server to go with a weaker ciphersuite than
the client-server together can support).

Stillman, et al.                                                [Page 8]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

TLS features:  ciphersuites are comprised of a triple (key exchange
algorithm WITH encryption algorithm, MAC algorithm).
The ciphersuite is negotiated between the client and server with the
server choosing the ciphersuite.  This negotiation allows new cipher
suites to be easily incorporated once they are standardized (example:
AES) and old ones to be dumped when they are shown to be
easily cracked.  Confidentiality is optional meaning the second
parameter can be null.  A cipher suite of NULL WITH NULL NULL
is deprecated (meaning don't do it).  It offers no security.  Once you
have successfully negotiated a TLS connection, TLS allows you
to resume sessions at a later time if the server is willing to do so.
How long a session can be around before it can not be resumed
is a matter of local policy.  Session resumption saves on performance
(CPU cycles) and handshake messages.

Assumptions:

(1) Each ENRP name server possesses a certificate (probably X.509 v3)
signed by a CA and an associated private key.  This allows the
server to validate itself as a legitimate ENRP server for the domain
foo.bar.com.  It will contain this domain name in the certificate
to allow the PE to check this against it's DNS inquiry.

(2) PEs may authenticate using TLS, SRP or some other authentication
protocol.  We could have each PE use TLS and supply a client certificate
but this might not scale well.  Therefore, I have suggested other
authentication mechanisms for PE to ENRP server.


3.1.1  PE - ENRP security

TLS is a client-server protocol and the client and server play different
roles.  In this scenario the PE functions as the TLS client and
ENRP functions as the TLS server.  We describe two different TLS
scenarios in this section to enforce PE - ENRP security.

For scenario A ENRP and PE use TLS for mutual authentication.  In
scenario B, ENRP servers authenticate themselves using TLS and PEs
authenticate themselves using some other unspecified authentication
mechanism.  This scenario will allow either TCP or SCTP as the transport
for TLS.  However, the current consensus is that the PEs should use SCTP
to communicate with the (ENRP) name server














Stillman, et al.                                                [Page 9]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

3.1.1.1 Scenario A -- TLS only

1) PE wants to register with a ENRP server.  Uses DNS to lookup
foo.bar.com ENRP server.
2) Establish a TLS connection with ENRP server.  Negotiate chiphersuite.
3) PE (client) Gets the ENRP server certificate as part of the TLS
protocol.
   a) Validate the signature;
   b) check expiration date;
   C) OCSP to check if the certificate has been revoked
        d) check the certificate contents name against the dns FQDM.
3) Get the client (PE) certificate as part of the TLS protocol.
   a) Validate the signature;
   b) check expiration date;
   C) OCSP to check if the certificate has been revoked
   d) check the cert contents against what? (This is a problem)
If any checks fail, send back error message (defined by TLS) and abort.
4) TLS session is now established with mutual authentication of the PE
and ENRP
server
5) PE Sends registration message with pool handle name using TLS session
6) ENRP will either authorize that PE to join that pool or ask a third
party (such as AAA) to authorize


3.1.1.2 Scenario B -- TLS plus alternate method for PE (client)
authentication

1) PE wants to register with a ENRP server.  Uses DNS to lookup
foo.bar.com ENRP server.
2) Establish a TLS session with ENRP server.  Negotiate chiphersuite.
3) Get the ENRP server cert as part of the TLS protocol.
   a) Validate the signature;
   b) check expiration date;
   C) OCSP to check if the cert has been revoked
If any checks fail, send back error message (defined by TLS) and abort.
4) TLS session is now established with au thentication of the ENRP
server
5) authenticate the client -- using the established TLS session, perform
client authentication mechanism using SRP, CHAP, etc.
6) If client auth fails, then the server terminates the TLS session
7) TLS session is now established with mutual authen of the PE and ENRP
server
8) PE Sends registration message with pool handle name using TLS session
9) ENRP will either authorize that PE to join that pool or ask a third
party to authorize.










Stillman, et al.                                               [Page 10]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

3.1.1.3 Open issues for ENRP-PE security

1) Order of authentication -- ENRP server first, PE client second OR PE
client first then ENRP server

Should we authenticate the PE (client) first and then set up the TLS
connection if the authentication of the PE is oK or vice versa?
ENRP - PE
Advantages: PE knows it is talking with a genuine name server quickly.
The other way would take longer for it to figure out it was
talking with a bogus ENRP.  The malicious ENRP could accept its
authentication credentials and only later it would find out the ENRP
server is not legitimate.  This would waste time.
PE - ENRP
Disadvantage - Alternatively, malicious ENRP server could just reject
its authen credentials and the PE would never find out that it
was talking with the "bad" ENRP server.

Conclusion: Authenticate ENRP server first, then the PE (client).

2) Do we allow authentication but no confidentiality in PE - ENRP
communications?
This is supported by TLS.  If we want Authentication but no
confidentiality, use TLS cipher suite:
key exchange algorithm WITH no encryption algorithm, MAC algorithm i.e.,
xxx WITH NULL, yyy


3.1.2 End user-ENRP security

For this scenario, we only need to authenticate the ENRP server.
Presumably, any end user can contact the name server.
In this scenario the end user functions as the TLS client and ENRP
functions as the TLS server.

1) Using ASAP the end user wants to resolve a name using ENRP server.
Uses DNS to lookup foo.bar.com ENRP server.
2) Establish a TLS session with ENRP server.  Negotiate chiphersuite.
3) Get the ENRP server cert as part of the TLS protocol.
   a) Validate the signature;
   b) check expiration date;
   C) OCSP to check if the cert has been revoked
If any checks fail, send back error message (defined by TLS) and abort.
4) TLS session is now established with authen of the ENRP server
Send request for name/address resolution using TLS session; get response
5) We can resume this session if local policy allows it (the ENRP server
policy, that is)


3.2 Scenarios for using IPsec with Rserpool

This scenario works with any transport protocol, although TCP or SCTP
are strongly recommended.



Stillman, et al.                                               [Page 11]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

3.2.1 Scenario A PU to ENRP server

To pre-establish one/more alternative IPSec security associations
(SA) that can be used should the primary SA become unusable (due to
server failure). No context sharing is done between SAs that are
terminated at the different servers.



    PU                          ENRP1      ENRP2
   ====                         ====       ====
    IP  SA1 -----------------SAa IP    /SA* IP
        SA2 --------------------------+

In this case PU has a separate SA with ENRP1 and ENRP2, call them SA1
and SA2.  In this case, the transport and RSERPOOL interactions, as well
as the application data is (a) protected (b) authenticated by IPsec.
I call the SAs by different names on the PE sides since its important
to note the problem is symmetric - you need association to be built
from each ENRP server back to each pool user.

The issue I see with this approach is the coordination between the
rserpool layer and the IPsec layer.  In particular, you need to
establish n SAs for each PU, where n is the number of ENRP servers.
This need not be done immediately:  There's the usual trade off of
eager/lazy processing.

Since this is network layer security, a break in the flow of
communication can be handled by the transport layer.  The security state
associated with the IPsec communication flow can be initialized and
established with a different flow - to a different agent, in the case
where RSERPOOL decides to redirect traffic.  The challenge is still at
the higher layers - doing application state context transfer.
However, since these name resolution messages don't rely on state, we
can just resend the messages in the event of failure of an ENRP server.





















Stillman, et al.                                               [Page 12]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

Time
  |   PU             ENRP1           ENRP2
  v  ====            =====           =====
       A!
       o-----
              \
               ----->
            B  -----o SA(x)
             /
   SA(x)<----   C
            <======>

       D!
       o------------------
                          \
                           ----->
                        E  -----o SA(y)
                          /
   SA(y)<-----------------   F
        <=======================>

A!  RSERPOOL decides to direct PU traffic to ENRP1
B   Establish SA and associated state between PU and ENRP1
    (like cypher block vectors)
C   Communicate between PU and ENRP1.  Stateful changes to SA(x) state
    continue to occur on both sides of the communication.
D!  RSERPOOL decides to redirect PU traffic to PE2
    Note that a distinct SA must be established between PU and ENRP2.
    The entire state between PU and ENRP1 can be tossed, or saved for
    future communication between PU and ENRP1.
E   Establish SA and associated state between PU and ENRP2
F   Communicate between PU and ENRP2.  Stateful changes to SA(y) state
    continue to occur on both sides of the association.

3.2.2 Scenario B - PE to ENRP server

There is no technical difference in Scenario A and B.  The names are
just changed, that is, substitute PE for PU.


















Stillman, et al.                                               [Page 13]

Internet Draft     Threats Introduced by Rserpool      12 December 2002

4. References:


[RFC2026] S. Bradner, "The Internet Standards Process -- Revision 3",
RFC 2026, October 1996.

[MIPv6 threats] draft-team-mobileip-mipv6-sec-reqts-00.txt, July, 2001,
work in progress.

[RFC3365] RFC 3365, Strong Security Requirements for IETF Standard
Protocols, August, 2002.

[ASAP] R. R. Stewart, Q. Xie: "Aggregate Server Access Protocol
(ASAP)", draft-ietf-rserpool-asap-05.txt, work in progress.

[ENRP] Q. Xie, R. R. Stewart "Endpoint Name Resolution Protocol",
draft-ietf-rserpool-enrp-04.txt, work in progress.

[SCTPIPsec] On the use of SCTP with IPsec,
draft-ietf-ipsec-sctp-04.txt, work in progress.

[TLS] TLS Version 1.0, RFC 2246.

[SCTPTLS] SCTP over TLS
draft-ietf-tsvwg-tls-over-sctp-00.txt, work in progress.


5. Acknowledgements

Thanks to the Rserpool security design team that provided valuable
comments.  Lyndon Ong, Randy Stewart, Qiaobing Xie, Michael Tuexen,
Sohrab Modi, Javier Pastor-Balbas, Xingang Guo, M. Piramanayagam,
Bernard Aboba and Dhooria Manoj.
























Stillman, et al.                                               [Page 14]

Internet Draft     Threats Introduced by Rserpool      12 December 2002
expires June 12, 2002

6. Author's Addresses

Ram Gopal
Nokia Research Center
5 Wayside Road
Burlington, MA 01803
USA
email: ram.gopal@nokia.com

Erik Guttman
Sun Microsystems
Eichhoelzelstr. 7
74915 Waibstadt
Germany
Email: Erik.Guttman@sun.com

Matt Holdrege
Sonus Networks
223 Ximeno Avenue
Long Beach, CA 90803
matt@sonusnet.com

Senthil Sengodan
Nokia Research Center
5 Wayside Road
Burlington, MA 01803
USA
email: Senthil.sengodan@nokia.com

Maureen Stillman
Nokia
35 Woodcrest Ave.
Ithaca, NY 14850
USA
email: maureen.stillman@nokia.com




















Stillman, et al.                                               [Page 15]


Html markup produced by rfcmarkup 1.109, available from https://tools.ietf.org/tools/rfcmarkup/