|
|
WS-I Security Scenarios Document Status: Working Group Draft Version: 0.15 Date: 14 February 2004 Editors: Mark Davis, Sarvega Bret Hartman, DataPower Chris Kaler, Microsoft Anthony Nadalin, IBM Jerry Schwarz, Oracle
|
Copyright
Copyright © 2004 by The Web Services-Interoperability Organization (WS-I) and Certain of its Members. All Rights Reserved.
Status of this Document
This document is a Working Group Draft; it has been accepted by the Working Group as reflecting the current state of discussions. It is a work in progress, and should not be considered authoritative or final; other documents may supersede this document.
Notice
The material contained herein is not a license, either expressly or impliedly, to any intellectual property owned or controlled by any of the authors or developers of this material or WS-I. The material contained herein is provided on an "AS IS" basis and to the maximum extent permitted by applicable law, this material is provided AS IS AND WITH ALL FAULTS, and the authors and developers of this material and WS-I hereby disclaim all other warranties and conditions, either express, implied or statutory, including, but not limited to, any (if any) implied warranties, duties or conditions of merchantability, of fitness for a particular purpose, of accuracy or completeness of responses, of results, of workmanlike effort, of lack of viruses, and of lack of negligence. ALSO, THERE IS NO WARRANTY OR CONDITION OF TITLE, QUIET ENJOYMENT, QUIET POSSESSION, CORRESPONDENCE TO DESCRIPTION OR NON-INFRINGEMENT WITH REGARD TO THIS MATERIAL.
IN NO EVENT WILL ANY AUTHOR OR DEVELOPER OF THIS MATERIAL OR WS-I BE LIABLE TO ANY OTHER PARTY FOR THE COST OF PROCURING SUBSTITUTE GOODS OR SERVICES, LOST PROFITS, LOSS OF USE, LOSS OF DATA, OR ANY INCIDENTAL, CONSEQUENTIAL, DIRECT, INDIRECT, OR SPECIAL DAMAGES WHETHER UNDER CONTRACT, TORT, WARRANTY, OR OTHERWISE, ARISING IN ANY WAY OUT OF THIS OR ANY OTHER AGREEMENT RELATING TO THIS MATERIAL, WHETHER OR NOT SUCH PARTY HAD ADVANCE NOTICE OF THE POSSIBILITY OF SUCH DAMAGES.
Feedback
The Web Services-Interoperability Organization (WS-I) would like to receive input, suggestions and other feedback ("Feedback") on this work from a wide variety of industry participants to improve its quality over time.
By sending email, or otherwise communicating with WS-I, you (on behalf of yourself if you are an individual, and your company if you are providing Feedback on behalf of the company) will be deemed to have granted to WS-I, the members of WS-I, and other parties that have access to your Feedback, a non-exclusive, non-transferable, worldwide, perpetual, irrevocable, royalty-free license to use, disclose, copy, license, modify, sublicense or otherwise distribute and exploit in any manner whatsoever the Feedback you provide regarding the work. You acknowledge that you have no expectation of confidentiality with respect to any Feedback you provide. You represent and warrant that you have rights to provide this Feedback, and if you are providing Feedback on behalf of a company, you represent and warrant that you have the rights to provide Feedback on behalf of your company. You also acknowledge that WS-I is not required to review, discuss, use, consider or in any way incorporate your Feedback into future versions of its work. If WS-I does incorporate some or all of your Feedback in a future version of the work, it may, but is not obligated to include your name (or, if you are identified as acting on behalf of your company, the name of your company) on a list of contributors to the work. If the foregoing is not acceptable to you and any company on whose behalf you are acting, please do not provide any Feedback.
Feedback on this document should be directed to secprofile_comment@ws-i.org.
Table of Contents
3.1 C-01: Peer Identification and Authentication
3.2 C-02: Data Origin Identification and Authentication
3.3.1 C-03A: Transport Data Integrity
3.3.2 C-03B: SOAP Message Integrity
3.4 C-04: Data Confidentiality
3.4.1 C-04A: Transport Data Confidentiality
3.4.2 C–04B: SOAP message confidentiality
5 Security Solutions and Mechanisms
5.1 Transport Layer Security Descriptions
5.1.3 Authentication by HTTP Service
5.1.4 Authentication by HTTP User Agent
5.2 SOAP Message Layer Security Descriptions
5.2.3 SOAP Sender Authentication
5.3 Combining Transport Layer and SOAP Message Layer Mechanisms
5.4 Transport and Message Layer Security Combinations
5.5 Security Considerations for Combinations
5.5 Security Considerations for Combinations
5.5.1 Transport Layer Security Solutions
5.5.2 SOAP Message Layer Security Solutions
5.5.3 Hybrid Security Solutions
6.1 Notation for Describing Scenarios
6.2 Conventions for Describing Security Requirements and Solutions
6.4 Generic Security Requirements
6.4.1 Requirement: Peer Authentication
6.4.2 Requirement: Origin Authentication
6.4.4 Requirement: Confidentiality
6.4.5 Requirement: Message Uniqueness
6.5.2 Scenario: Synchronous Request/Response
7.1.2 C-06: Credentials Issuance
This document defines the requirements for and scope of the WS-I Basic Security Profile. The document is aimed at Web Services architects and developers who are examining the security aspects of the Web Services they are designing/developing.
This document:
This document assumes that the reader has at least a basic background in security technologies such as SSL/TLS, XML encryption and digital signatures, and OASIS Web Services Security.
This document does not deal with security aspects of attaching material to SOAP messages as described in the WS-I Attachment Profile 1.0. A final version of this document will include this material.
This section defines vocabulary that will be used to refer to the various entities and concepts in this document.
The following terms are used to describe certain entities.
An alternative is to use “entity” as the most abstract term and reserve “participant” for the SOAP nodes that are parts of scenarios. However, “entity” sounds a bit stilted. Note that a SOAP node is a participant.
Communication channels are inevitably layered. When, as in this document, it is necessary to discuss the interaction between layers some care is required to distinguish between events and messages at one level from those that occur at a lower level. In general what appears to be an atomic action, such as message transmission, at one level will have a more complicated structure at a lower level.
We are primarily interested in transmission of SOAP messages and the participants in the transmission. However in some cases we are also interested in non-SOAP messages.
Message: Protocol elements that are exchanged, usually over a network, to affect a Web service (i.e. SOAP/HTTP messages)
Normally HTTP and SSL/TLS would be considered separate layers. Consolidating them and lower layers compresses the stack. But it is convenient to treat HTTP, SSL/TLS and lower layers together.
SOAP 1.2 defines the following terms:
We adopt these terms with the understanding that we will apply them to SOAP 1.1 messages rather than SOAP 1.2 messages. We will not use any terms that refer specifically to SOAP 1.2 features that are not present in SOAP 1.1
The participants in a message event are referred to as
In most contexts it is not necessary to distinguish the various layers in the communication, however when it is necessary to do so “sender” or “receiver” may be modified by the protocol involved, so that “SOAP sender” and “HTTP receiver” can be used.
The use of “sender” and “receiver” is so natural that it would be hard to avoid them even if they weren’t part of the official glossary.
This section identifies potential security challenges that scenario may want to address. The following subsections characterize the identified security challenges with the following attributes:
Definitions:
Peer entity authentication: The corroboration that a peer entity in an association is the one claimed.
Identification: An act or process that presents an identifier to a system so that the system can recognize a system entity and distinguish it from other entities.
Explanation: Any relationship between entities can be considered an “association” for purposes of this definition. For example, it does not require that the two entities directly communicate with each other.
Although the term “authentication” is sometimes used to include both the presentation and the corroboration of an identifier this document uses “authentication” in the narrower sense defined here.
A participant may convey information to another participant to establish identity in conjunction with the use of techniques to corroborate that information. The two SOAP participants are not necessarily directly connected by a single hop, for example the participants might be the initial SOAP sender and a second SOAP intermediary. Depending on application requirements (security policy) it may be reasonable to authenticate the sender, receiver or to use mutual authentication.
NOTE:
It is important for a relying party to ensure the correctness of the identification associated with authentication. For example, in using SSL/TLS a server may present an X.509 certificate to associate identity information with a public key and use the corresponding private key to prove possession of the private key. A relying party should not only rely on the authentication technology, but should also ensure that the information associated with the authentication is correct, thus authorizing further processing based on that information. This may include steps such as ensuring that the HTTP request domain name corresponds to the server certificate name and performing certificate validation. Such care is necessary in light of man-in-the-middle, DNS or TCP/IP attacks (T-05) where authentication may work technically but does not corroborate the correct party. Authorization is important but not addressed in this document.
Candidate technology:
Threat association:
T-04, T-05, T-06, T-07, T-08, T(OOS)-01, T(OOS)-03, T(OOS)-04, T(OOS)-08, T(OOS)-13
Definitions:
Data origin authentication: The corroboration that the source of data received is as claimed.
Identification: An act or process that presents an identifier to a system so that the system can recognize a system entity and distinguish it from other entities.
Explanation: The provision and authentication of a declaration, carried in a web service message that some entity vouches for certain parts of the message. (Here, it is intended that attachments be considered “parts” of a message.) Note that it is possible that more than one entity might be involved in vouching for message parts. Also note that it is application-dependent as to how it is determined who initially created the message, as the message originator might be independent of, or hidden behind a vouching entity. This mechanism does not provide for the Authentication of the Destination prior to transmission of application data. However, the encryption of the data with a key only known to the legitimate destination can effectively serve as an implicit form of Destination Authentication if that is required.
This of course does not prevent the impersonation of the legitimate destination for the purposes of Denial of Service.
Candidate technology:
Threat association:
T-04, T-05, T-06, T-07, T-08, T(OOS)-01, T(OOS)-03, T(OOS)-04, T(OOS)-08), T(OOS)-13
Definition: Data integrity: The property that data has not been changed, destroyed, or lost in an unauthorized or accidental manner (see [RFC 2828]).
Explanation: Data in a web services context is taken to mean a SOAP message or portions of a SOAP message, including one or more SOAP header, body, or attachment parts. Although data integrity is concerned with allowing a recipient of data to detect changes, whether accidental or malicious, data origin authentication mechanisms are required in conjunction with data integrity mechanisms in order to protect against active substitution and forgery attacks. When only providing integrity for portions of content, care must be taken to protect against subtle attacks, especially when a message is targeted at SOAP intermediaries as well as an ultimate receiver.
Note that the term “Integrity” is generally used differently in the field of information management to mean that the data is correct, proper, accurate, and consistent with other data or the real world. In this sense it usually implies that there are well-regulated procedures of creating, modifying and deleting the data. Here we are using “Integrity” in the security sense of not being altered without detection of such alteration even when under active attack.
Threat association: T-01, T-02. Additional threats associated with sub-categories of data integrity are listed below. Note that when used in conjunction with data origin authentication T-04, T-05 and T-06 are addressed.
Definition:
Transport Data Integrity: Data integrity provided by the protocol layer that SOAP messages are bound to, e.g. HTTP secured by SSL/TLS (HTTPS).
Explanation: Transport integrity is applied to the entire SOAP message and may also include underlying protocol layers. For example, with HTTPS the HTTP message is also protected. Such transport layer security is “transient” in that the integrity is only effective while the transport session exists. Transport integrity is not appropriate for end-to-end security (from SOAP initiator to ultimate receiver) when SOAP intermediaries are present, since SOAP processing rules allow intermediaries to make changes to the SOAP message, and since transport protection is not in effect during intermediary processing.
Candidate technology:
Additional Threat Associations: T-09, T(OOS)-10,
Definition:
Soap Message Integrity: Data integrity applied at the SOAP Messaging layer in a manner that allows SOAP processing rules to be followed.
Explanation: SOAP message data integrity is for a web service message that may be processed by SOAP intermediaries and may exist for extended periods of time at intermediary and/or ultimate receiver SOAP nodes before being processed. The intention is to protect message data even when not in transit, such as before processing is completed. An example is a SOAP message waiting at a SOAP node for aggregation with other content yet to be processed. Transport integrity is inappropriate for such cases since it terminates with the transport session.
SOAP message integrity should be applied to a SOAP message in a manner that enables processing by SOAP intermediaries, which suggests that integrity protecting a combination of SOAP header blocks and the body is preferable to protecting the entire SOAP envelope element or the entire SOAP header element. Protection may also include SOAP attachments.
Candidate technologies:
·
XML Signatures as profiled in the OASIS SOAP Message Security
specification.
Note that keys may be conveyed out of band or with the message using a SOAP
Message Security token profile, including (but not limited to) Username tokens
(for derived keys), X.509, Kerberos tokens or others.
· XML Signatures with MIME, not in the context of SOAP Message Security (out of scope)
· CMS (pkcs7) with MIME
XML Signatures not in the context of SOAP Message Security headers can be used by applications, but that use is not addressed in this document.
Definition: Data confidentiality: The property that information is not made available or disclosed to unauthorized individuals, entities, or processes [i.e. to any unauthorized system entity] (RFC 2828).
Explanation: The property that eavesdroppers or other unauthorized parties cannot view confidential message content. Typically this is achieved with encryption. Note that confidentiality is a distinct concept from privacy, so in the definition "disclosure" refers to the ability to view or eavesdrop the information when transferred or processed. Confidentiality techniques may be used as one aspect of maintaining privacy, however.
Threat Associations: T-03, T(OOS)-10
Disclosure related attacks as well as attacks that reduce the confidentiality strength (e.g. man-in-the-middle SSL/TLS ciphersuite attacks) are relevant.
Definition: Data confidentiality provided by the protocol layers that SOAP messages are bound to in a transport protocol stack specific manner. An example is HTTP secured by SSL/TLS (HTTPS).
Explanation: Data confidentiality is applied to the entirety of the SOAP message as well as possibly other protocol layers (e.g. HTTP when SSL/TLS is in use). With end-to-end confidentiality between the initial SOAP sender and the ultimate receiver this prevents the use of SOAP intermediaries.
Candidate technology:
Additional threat associations:
none.
Definition: Data confidentiality applied at the SOAP messaging layer in a manner that allows SOAP processing rules to be followed.
Explanation: SOAP message confidentiality supports the confidentiality requirements unique to SOAP messaging, including:
This may include extended periods of time that the SOAP message is queued at an intermediary or ultimate receiver before being processed. An example is a SOAP message waiting at a SOAP node for aggregation with other content yet to be processed.
Transport confidentiality is generally inappropriate for these requirements since it terminates with the transport session.
In order for SOAP message confidentiality to be applied to a SOAP message in a manner that enables processing by SOAP intermediaries, a combination of SOAP header blocks, body blocks and attachments is appropriate, but the soap:Envelope, soap:Header and soap:Body elements must be visible to all parties and should not be encrypted. The SOAP message must also remain well-formed XML.
Candidate technologies:
Additional threat associations: none
Definition: the ability to insure that a specific message is not resubmitted for processing.
Explanation: Attacker could resend all or selective parts of a message causing undesirable side effects. For example, an attacker sending the same valid message moving money from one bank account to another bank account. The original message request is valid, but not its replay. Additionally, sending the same valid message is frequently used in many denial-of-service attacks. While an application solution against replay attacks may utilize message ordering and reliable message delivery mechanisms, this security challenge makes no attempts to address these issues.
Candidate technologies:
· At the transport layer, using SSL/TLS between the node generating the request and the node insuring for downstream nodes that this is a unique request.
· At the message layer, the sending and receiving SOAP nodes must do a combination of different things. The sender must sign SOAP message header nonce, creation time[, expiration time] and optional user data. This user data may include critical transactional information and service identification elements. The transactional data protects the actual user request. The optional service identification elements protect the replay of the signature to another service that utilizes the same message data. The receiving node must verify the signature and check that the creation time is not stale. Lastly, it must compare the received nonce with a cache of previously receive nonces. This cache of nonces must be maintained until the associated expiration time or the creation time plus a hard-coded delta has expired. Note: when multiple servers are performing this functionality, some mechanism must be implemented to create a functional global cache across all these systems.
Threat association: T-08, T-09, T-10.
This section details a list of traditional security threats. Note that in many cases the threats overlap. That is particular attacks may represent threats in several categories.
|
ID |
Name |
Description |
|
T-01 |
Message Alteration |
The message information is altered by inserting, removing or otherwise modifying information created by the originator of the information and mistaken by the receiver as being the originator’s intention. There is not necessarily a one to one correspondence between message information and the message bits due to canonicalization and related transformation mechanisms. |
|
T-02 |
Attachment Alteration |
|
|
T-03 |
Confidentiality |
Information within the message is viewable by unintended and unauthorized participants. (e.g. a credit card number is obtained). |
|
T-04 |
Falsified Messages |
Fake messages are constructed and sent to a receiver who believes them to have come from a party other than the sender. For example, Alice sends a message to Bob. Mal copies some (or all of) it and uses that in a message sent to Bob who believes this new action was initiated by Alice. This overlaps with T-01and T-02. The principle is that there is generally little value to saying a message has not been modified since it was sent unless we know who sent it. |
|
T-05 |
Man in the Middle |
A party poses as the other participant to the real sender and receiver in order to fool both participants (e.g. the attacker is able to downgrade the level of cryptography used to secure the message). The term “Man in the Middle” is applied to a wide variety of attacks that have little in common except for their topology. Potential designs have to be closely examined on a case-by-case basis for susceptibility to anything a third party might do. |
|
T-06 |
Principal Spoofing |
A message is sent which appears to be from another principal (e.g. Alice sends a message which appears as though it is from Bob). This is a variation on T-04. |
|
T-07 |
Forged claims |
A message is sent in which the security claims are forged in an effort to gain access to otherwise unauthorized information (e.g. A security token is used which wasn't really issued by the specified authority). The methods of attack and prevention here are essentially the same as T-01 and T-02. |
|
T-08 |
Replay of Message Parts |
|
|
T-09 |
Replay |
|
|
T-10 |
Denial of Service |
Amplifier Attack: attacker does a small amount of work and forces system under attack to do a large amount of work. This is an important issue in design and perhaps profiling in some cases. |
Table 1: Threats
Additional information on security threats can be found in the following titles:
In this section, we provide a high-level description of security solutions, which are defined in terms of security layers that address the SOAP message security challenges in Section 3. We then define the specific security mechanisms and associated countermeasures that are addressed by the Security Profiles.
Mechanisms to address security challenges may be applied at different communication layers and possibly in combination. The primary concerns of this document are the SOAP and transport layers. Within the transport layer the focus is primarily on HTTP and HTTPS. Combinations of security mechanisms in the layers may be applied to satisfy different security requirements.
This document focuses on scenarios for transport and SOAP Layer security. Users may implement their own data (payload) layer security, but data layer security is not addressed explicitly in this document.
Transport and SOAP security layers can be configured to address a variety of security requirements. These variations are enumerated later in this section. We define abstract security functions that may be used to address the various security threats that we previously described in Section 4.
The protocol layers that provide transport for the SOAP Messaging protocol (transport layer) may be used to provide security services to meet application or SOAP Messaging security requirements. This may be done in combination with SOAP message Security mechanisms or independently. This section focuses on the transport mechanisms only. These mechanisms provide integrity and/or confidentiality for HTTP messages, thus protecting SOAP messages with or without attachments.
Because the only transport mechanism within the scope of this document is HTTP (optionally over SSL/TLS) we assume that each SOAP node has an associated HTTP node, which might be a part of the SOAP node or might be a distinct entity. We also assume that SOAP messages between nodes are carried on HTTP messages between their associated HTTP nodes. Communication between a SOAP node and its associated HTTP node is regarded as internal to a platform and we make no assumptions about its nature or the information transferred other than
Note in particular that we do not assume any communication between the associated HTTP and SOAP nodes with regards to the certificates used to establish a TLS/SSL connection.
In what follows when a word or phrase such as “N” refers to a specific SOAP node we use the notation “N-HTTP” to refer to its associated HTTP node.
Integrity may be provided for an entire SOAP message using the transport layer. When SSL/TLS is used in conjunction with HTTP (HTTPS), the entire HTTP message, including the start-line (e.g. POST), HTTP headers, and body receives integrity protection. This SOAP message conveyed in the HTTP body is also protected. This integrity is only in effect for the duration of the HTTP session and provides no protection for SOAP messages once received (and possibly queued by the web service consumer or requestor). Note that integrity is provided for the entire SOAP message – partial integrity is not possible with this mechanism. This mechanism is not suitable for end-end SOAP message integrity in the presence of SOAP intermediaries.
The basic operation of this mechanism is as follows:
Note that the quality of SSL/TLS integrity protection depends on an adequate SSL/TLS ciphersuite and key length being selected. Care must be taken in selection of ciphersuites and key lengths to prevent downgrade attacks. Options with inadequate security should not be offered even if they are supported in the code.
Confidentiality may be provided for an entire SOAP message using the transport layer. When SSL/TLS is used in conjunction with HTTP (HTTPS), the entire HTTP message including HTTP headers is protected as well. This confidentiality is only in effect for the duration of the HTTP session and provides no protection for SOAP messages once received (and possibly queued by the web service consumer or requestor). Confidentiality is applied to the entire SOAP message, partial confidentiality is not possible, making this unsuitable for SOAP messages to be conveyed through SOAP topologies involving SOAP intermediaries.
The basic operation of this mechanism is the same as that using transport layer to provide integrity. [Section 5.1.1
Note that the presence and quality of SSL/TLS integrity protection depends on an adequate SSL/TLS ciphersuite and key length being selected. Care must be taken in selection of ciphersuites and key lengths to prevent downgrade attacks. Options with inadequate security should not be offered even if they are supported in the code.
A SOAP node A whose associated HTTP node initiates a connection from SOAP node B’s associated HTTP node may authenticate B using transport layer mechanisms such as SSL/TLS. In the SSL/TLS case the authentication consists of a server X.509 certificate combined with a proof of private key possession as part of the SSL/TLS protocol. In addition, some clients may perform additional checks such as comparing the service URL domain name against the certificate distinguished name, for example, to attempt to detect certificate substitution attacks. Finally, relying parties should perform a certificate validation check to ensure that the certificate was not revoked, either due to private key compromise or other reasons before relying on the validity of the authentication information.
The basic operation of the mechanism is as follows:
1. HTTP node associated with A initiates HTTPS connection to HTTP node associated with B.
2. As part of establishing SSL/TLS session, B’s HTTP node authenticates to A’s HTTP node
3. SOAP messages are conveyed from A to B, potentially SOAP message or fault is conveyed in HTTP response
4. HTTP and SSL/TLS session is terminated
Note that the authentication is for the session and that by default there is no lasting record or association of the authentication action with the SOAP message.
A SOAP node A whose associated HTTP node initiates a connection to SOAP node B’s associated HTTP node may authenticate to SOAP node B. If B’s HTTP node also authenticates to A’s HTTP node it is said to be mutual authentication.
Note that a web service provider might authenticate at the transport layer and the web service consumer at the SOAP messaging layer, depending on the desired authentication properties.
An HTTP user agent authentication may be:
HTTP Basic and Digest authentication mechanisms are outlined in [RFC 2617],
HTTPS session is closed
Note that B-HTTP must request authentication explicitly. The SOAP message may be POSTed twice – once in the original POST that results in an HTTP response requesting authentication and then in the request that conveys the authentication information in the header. This could be an issue for large SOAP messages.
Adequate protection against replay attacks is required with HTTP authentication and POSTs as noted by RFC 2617. HTTPS confidentiality requires appropriate ciphersuites and protection against downgrade attacks.
Using HTTP with Digest authentication provides no real benefits in terms of authentication over Basic authentication, although with the proper cipher suites it can provide integrity.
HTTP Basic or Digest authentication performed as part of HTTP session that includes SOAP message request POST.
Despite the risk of insider attack (most attacks are insider attacks) HTTP authentication without HTTPS may be appropriate within an enterprise or other secured environments. Protection against replay attacks is required as noted by RFC 2617.
Attributes may be conveyed in HTTP header fields [RFC 2616]. This may require integrity and/or confidentiality protection using HTTPS, depending on application requirements.
Attributes may also be conveyed in the HTTPS client X.509v3 certificate through the use of certificate extensions, although this may not be interoperable. See PKIX RFC 3280.
The preceding transport layer security mechanisms may be combined with each other as needed. The following table attempts to identify the combinations that we believe are significant with a unique tag that we will use in later sections.
|
Challenge Supported |
Transport Layer Technologies being Utilized |
Tag[1] |
Comment |
|
|
Integrity |
SSL/TLS |
BISP1 |
|
|
|
Confidentiality |
SSL/TLS |
Assuming that cipher suites NULL-SHA or NULL-MD5 are not being supported because these suites do support encryption. |
||
|
Provider (server) Authentication |
SSL/TLS |
|
Assume X.509 certificates being used to identify consumer and provider with mapping to trusted root CA. |
|
|
Consumer (client) Authentication |
SSL/TLS[2] with client authentication |
BC1 |
||
|
HTTP Basic |
BC2 |
|
||
|
HTTP Digest |
BC3 |
|
||
|
HTTP Attributes |
BC4 |
|
||
|
SSL/TLS |
HTTP Basic |
BC5 |
This assumes that BISP1 is also supported. Additionally, assumes cipher suites NULL-SHA & NULL-MD5 not supported, i.e., protection against downgrade attacks. |
|
|
HTTP Digest |
||||
Table 2: Transport Level Security Options
The intention is for an application developer to select one or more solutions that address the relevant security challenges. For example, if consumer authentication is required then any one of the BCx solutions would meet this need.
As indicated, a single solution may meet multiple security challenges. For example, assuming cipher suites NULL-SHA or NULL-MD5 are not supported, using SSL/TLS will ensure transport layer integrity, confidentiality and provider authentication.
Security services may be provided at the SOAP Messaging protocol layer using the SOAP Message Security specification from the OASIS SOAP Message Security technical committee in conjunction with token specifications developed in that committee. These security mechanisms may be combined with the transport layer security mechanisms discussed above.
Integrity may be provided to a portion or combination of SOAP message payload and header blocks using XML Digital Signature as outlined in the SOAP Message Security specification. Such integrity has the advantage that it remains with the SOAP message beyond an HTTPS session, suitable for providing end-end integrity despite SOAP intermediaries, when used properly.
[The mechanism for providing integrity for attachments at the SOAP level must be determined. It will take into account basic profile group’s work on attachments]]
Confidentiality may be provided to portions or some number of SOAP Message body or header block element or element content using XML Encryption as outlined in the SOAP Message Security specification. Note that encryption must not be applied so that SOAP message processing cannot be performed. Note also that the SOAP Message Security specification is silent about SOAP attachment confidentiality.
[The mechanism for providing integrity for attachments at the SOAP level must be determined. It will take into account basic profile group’s work on attachments]]
SOAP message confidentiality protection has the advantage that it remains with the SOAP message beyond an HTTPS session, and is suitable for providing end-end confidentiality despite SOAP intermediaries when used properly.
A SOAP Sender (either an initial SOAP sender or a SOAP intermediary) may provide authentication for one or more SOAP receivers by including one or more appropriate SOAP Message security tokens in security headers targeted at the receiver roles may be used in combination with XML Signatures as profiled by SOAP Message Security to provide confirmation of the token claims and to bind the claims to the message.
Note that in a SOAP message from a web service consumer to a web service provider, SOAP sender authentication authenticates the consumer. In a SOAP message from a web service provider to a web service consumer (such as conveyed in an HTTP response in a request-response MEP) then SOAP sender authentication authenticates the provider to the consumer. SOAP receiver authentication as such does not make sense given a one-way message.
Attributes may be conveyed in application specific SOAP Message Security XML or Binary security tokens (SOAP Message Security extension points), or SOAP Message Security SAML Tokens conveying attribute assertions to give two examples.
This functionality is build upon the message integrity mechanisms, digital signatures, referred to in Section 5.2.1 being applied to several fields with special semantics and a number of things outside the actual message exchange. Depending upon the type of security token being utilized by the application to authenticate the sender, different elements in the message may be utilized. All the solutions are built upon the following key types of information being present in the sender message:
Unique message identifier: this element is used to uniquely identify the message. No two messages should ever have this value. While this data could be consequently assigned sequence numbers or non-random data, experience has shown that such practices allow for session hijacking unless the associated authentication mechanisms are very strong. Using true random values for the message identifier is best practice because an attacker can not effectively guess what message identifier someone is using or may use. [Some form of this element must be present in any solution]
Timestamp: a time that bounds the associated message identifier lifetime. Without this value, the consuming entity would potentially have to maintain data to track all message identifiers that it has ever processed. For some restrictive environments, e.g., single source, this timestamp can be used for the unique message identifier. In general, this is not true. The bigger issue with the timestamp is that the sending and receiving systems must be loosely time synchronized so that the receiving system does not have to maintain an ever-increasing database of processed message identifiers. With the availability of clock synchronization protocols and the receiver ability to control the size of the time window, applications can control the degree of time synchronization needed. While careful date/time set up could work if an application supports a large time window, e.g., 5-10 minutes, in general some form of clock synchronization is really required for effective operation. [Some form of this element must be present in any solution]
Optional Application Restrictions: These elements allow an application to prevent the replay of the preceding elements to different receiving systems. For example, to prevent a valid message identifier and application message data from being sent to a different receiving system and being processed, the domain of the target service that this request is intended for could be included within the data to be signed. [Application dependent data with associate application semantic checking.]
Of the different types of security tokens that our profile is committed to address, i.e., X.509 certificates, username, Kerberos, only username tokens currently have elements defined that map to the unique message identifier and timestamp element just described.
As will become very apparent, no security token profile and other standards will deliver a fully operation solution to the message uniqueness challenge at the SOAP message layer.
In particular, the username token profile defines the following elements that the sending system must populate when building a message uniqueness solution:
Nonce: a random value that the sender generates and uses as the unique message identifier. [The nonce is a recommend element in OASIS Username Token Profile that can be overloaded to serve as the unique message identifier. When used for replay prevention, this element must be present. When used for this purpose, it must be large enough to ensure that multiple simultaneous requesters do not generate the same nonce value causing a fail positive.]
Creation Time: the time that the associated nonce was created. [The creation time is a recommend element in OASIS Username Token Profile that can be overloaded to serve as the timestamp. When used for replay prevention, this element or expiration time element must be present.]
Expiration Time: the time when the associated nonce is no longer valid to be used. [The expiration time is an optional element in OASIS Username Token Profile that can be overloaded to serve as the timestamp. If not present, then the receiving system must add an internally configured delta time to the creation time element.]
Additionally, the preceding required and optional data along with the username must be signed by the sender so that the receiving system can ensure that none of the preceding elements has been modified by an attacker. This comes with the unstated assumption that the signing key (some function of the associated password) is known only to the sender and receiver as either an out-of-band shared secret or encrypted. Otherwise, the receiver can not authenticate the sender is who then say they are.
On the receiving system, the receiver must perform the following actions:
The OASIS X.509 Certificate and Kerberos Profiles do not have the required elements for acting as message identifier thus requiring application developer to define proprietary elements to address these needs, i.e., outside the scope of these token profile.
There are other token types being worked on that contain nonce and timestamp elements. However, their detail characteristics may prohibit them for being used to prevent replay attacks.
The preceding message layer security mechanisms may be combined with each other as needed. The following table attempts to identify the combinations that we believe are significant with a unique tag that we will use in later sections.
|
Challenge Supported |
Message Layer Technologies being Utilized |
Tag[3] |
Comment |
|
|
Integrity |
XML Digital Signature |
SI1 |
|
|
|
Confidentiality |
XML Encryption |
SC1 |
|
|
|
SOAP Sender Authentication |
XML Encryption |
username & [password|digest] |
SA1 |
Without the ability to encrypt password/ digest, sender open to man-in-middle stealing password/digest and reusing it. |
|
username & [password|digest] |
SA2 |
SOAP Attributes |
||
|
X.509 Certificate |
SA3 |
|||
|
Kerberos Token[4] |
SA4 |
|||
Table 3: SOAP Message Level Security Options
The intention is for an application developer to select one or more solutions that address the relevant security challenges. For example, if SOAP sender authentication is required then any one of the SAx solutions would meet this need.
Missing from this table is SOAP receiver authentication. Receiver message layer authentication can only be supported by a response message in which the role of the sender and receiver has been exchanged, i.e., the sender is the provider.
As noted above security services may be provided at either or both the transport layer and the SOAP message layer. The choice often depends on application requirements, based on answers to questions such as:
Special cases are noted in the sections above where additional mechanisms are required to ensure security. In general minimizing combinations while following recommended security practices for the security technologies should reduce risks.
This section describes a selected subset of common security scenarios and identifies potential solutions for various security requirements. The security requirements vary from simple to complex depending upon the mechanisms selected and the underlying need. This approach allows the users to select a specific security scenario and implementation mechanisms that best meet their needs.
There are three basic categories of implementation solutions:
· transport layer,
· SOAP message layer
· hybrid that combines mechanisms from transport and SOAP message layers.
Figure 1 attempts to depict the potential solution space. It is organized with
transport only mechanism on the left side of the figure and SOAP message
mechanisms on the right side. Hybrid solutions occupy the space in the middle.
This figure is not bound to any specific scenario. Different scenarios may be
able to only support a subset of implementations, e.g., one-way scenario can
not support SOAP mutual authentication because there is no SOAP response
message.
Additionally,
Figure 1 is organized from top to bottom to go from no security to increasing
complex security solutions.
Figure 1 Common Security Solutions Hierarchy
The eleven solutions identified
in
Figure 1are a much smaller set than all possibilities of combined security
solutions suggested by Table 2 on page 21 and Table 3 on page 26. A basic
question is what approach or reasoning was used to reduce the numbers?
Starting with the four transport entries, the two left solutions: BISP1 and
BISP1:BC1, are simply SSL/TLS with and without client authentication. The BC2
| BC3 | BC4 solution is all that can be done with only using HTTP. The last
solution is simply the merging/ enhancement of the SSL/TLS solutions and the
pure HTTP solution. Remember that these two transport level mechanisms: HTTP
and SSL/TLS, only work between HTTP/TCP level nodes. No SOAP intermediaries
are allowed. If multiple HTTP or higher nodes are encountered, then multiple
instances of the transport layer mechanisms between all communication HTTP
nodes may need to be used. Additionally, each intermediary has full access to
all the data passing by to look at or alter, i.e., no way to insure the
integrity or confidentiality within the HTTP/TCP intermediaries.
Moving to pure SOAP message solutions, the top solution is identifier of the sender, without integrity or confidentiality. The next two solutions are message level integrity or confidentiality along with the identification of who the sender (signer/encryptor) is. The assumption is that usually it does not matter if a message is unchanged unless you know who signed (originated) the data. Similarly, the secrecy of a message is not important if you can not also insure that source of the secret information. The two SI1:SC1:(SA1|SA2|SA3) solutions utilize all the SOAP message level mechanisms: Integrity, Confidentiality and Sender Authentication, for one-way and two-way MEP, respectively. Unlike the transport level mechanisms, the SOAP message level mechanisms allow integrity, confidentiality and sender authentication of all or part of a message to occur between any SOAP nodes, not just the ultimate sender and receiver.
Lastly, there is a single hybrid case supported. This hybrid case uses SSL/TLS to insure the confidentiality and integrity of the entire SOAP message data. The usage of SSL/TLS is a simple solution that also protects against various types of man-in-the-middle replay attacks that would be more complex and expensive to protect against via pure SOAP message level mechanisms. The bottom line is that this solution allows stricter security requirements to be imposed between a single pair of sender and receiver HTTP/TCP nodes than between other nodes in the message exchange. This is just the logical extension that each set of nodes in a complex message exchange may have different security requirements. Transport level mechanisms addresses only security requirements between connected HTTP/TCP nodes, while SOAP message level mechanisms addresses security requirements between any nodes in a message exchange. Each mechanism can be used multiple times for each combination of nodes that has specific security needs.
In this section we provide an overview of the issues to consider when deploying the combinations of transport and message layer security mechanisms defined in Section 5.4. For each of the common security solutions previously shown in Figure 1, we summarize the properties of the solution, threats addressed, and limitations.
These considerations may be used as a guide to select an appropriate security solution for many Web Services application deployments. By matching up a particular application’s security requirements against the solutions in this list, it should be possible in most cases to select an optimal combination of transport and/or message layer security mechanisms for that application.
The solutions in this subsection are based solely on transport layer security mechanisms.
T-06
This solution has the following properties:
T-01, T-02, T-03
This solution has the following properties:
T-01, T-02, T-03, T-04, T-05, T-06, T-07, T-08, T-09
This solution has the following properties:
T-01, T-02, T-03, T-04, T-06, T-07, T-08, T-09
The solutions in this subsection are based solely on SOAP message layer security mechanisms.
This solution has the following properties:
T-06
This solution has the following properties:
T-01, T-02, T-06
This solution has the following properties:
T-03, T-06
This solution has the following properties:
T-01, T-02, T-03, T-06, T-07
This solution has the following properties:
T-01, T-02, T-03, T-06, T-07
The solutions in this subsection are based on a combination of transport and SOAP message layer security mechanisms.
This solution has the following properties:
T-01, T-02, T-03, T-04, T-05, T-06, T-07, T-08, T-09
This solution has the following properties:
T-01, T-02, T-03, T-04, T-05, T-06, T-07, T-08, T-09
This section contains descriptions of scenarios, security requirements that might be imposed by applications using those scenarios and ways to satisfy those requirements (called solutions).
The content of a scenario and the conventions used to describe them are as follows.
Nk = B
S: A ® B
Indicates A-HTTP is the HTTP User Agent that initiates session S talking to HTTP Service B-HTTP. Sessions might be created during the scenario or might have existed before the scenario begins.
M: A ® B ®... ® Z
indicates that the scenario includes a SOAP message that travels on the indicated SOAP Path. Nodes in this description of a SOAP message are said to be prior to Nodes to their right and latter than Nodes to their left in the SOAP message path.
H: A ® B (Session S, Message M)
indicates that H is an HTTP Message that is sent by A-HTTP to B-HTTP as part of transmission of SOAP message M. Nodes A and B are said to be adjacent (on Message M). Whether H is an HTTP request or response depends on whether A or B initiated HTTP Session S. If it is a response, the Hop to which it is a response will be indicated.
H: A ® B (Session S, Message M, Response to R)
The order in which the Hops are listed is the order in which the HTTP messages are sent.
[These notations do not take into account attachments. If necessary they will be modified to do so when attachment considerations are added to this document.]
The description of a security requirement contains:
In describing the scenarios, requirements and solutions, the following phrases are used.
This section contains security requirements that may be imposed by applications that use the scenarios The requirements in this section are generic to all scenarios and might apply to any uses of SOAP Messaging.
This section only presents security requirements for which solutions are available within the profiled technologies. Other security requirements that might exist must be addressed by application level mechanisms.
A SOAP node A must be able to authenticate to any SOAP node B.
Threats: T-05, T-06
Challenges: C-01
Security solutions:
The following solution may be used to provide authentication of A to B when A is prior to B on a SOAP message Path.
a) SOAP Sender Authentication (Section 5.2.3) of the SOAP message.
The following solutions may only be used to provide authentication of A to B when A-HTTP initiates a session to B-HTTP.
b) HTTPS X.509 Client Authentication (Section 5.1.4.1
c) HTTP Basic or Digest Authentication with HTTPS Confidentiality (Reference 5.1.4.2)
d) HTTP Basic of Digest Authentication in the Clear (Reference 5.1.4.3)
The following solution may only be used to provide authentication of B to A when A-HTTP initiates a session to B-HTTP.
e) HTTPS X.509 Server Authentication (Section 5.1.4.1)
Solutions (c) and (d) do not address T-05 (man in the middle)
A party in possession of a SOAP node’s public key must be able to prove that signed SOAP message content was produced by that SOAP node.
Threats: T-05, T-06, T(OOS)-13
Challenges: C-01, C-05
Security solution:
a) Digital Signature on Message. SOAP Message Layer Integrity (Section 5.2.1)
A SOAP node B must be able to detect alteration of content supplied by a SOAP node A
Threats: T-01, T-02
Challenges: C-03
Security solution:
The following solution may be used to provide integrity for any content supplied by SOAP node A.
a) SOAP Layer Integrity (Section 5.2.1
The following solution may be used to provide integrity for any content while it is in transit on a Hop to or from A.
b) Transport Layer Integrity (Section 5.1.1
A SOAP node B must be able to exclusively access confidential content supplied by a SOAP node A and intended for SOAP node B.
Threats: T-03
Challenges: C-04
Security solution:
The following solution may be used to provide confidentiality of any content supplied by Node A
a) SOAP Layer Confidentiality (Section 5.2.2
The following solution may be used to provide confidentiality for content while in transit from A-HTTP to B-HTTP
b) Transport Layer Confidentiality (Section 5.1.2)
A SOAP node B must be able to detect that a previous received message or part of a previous message from SOAP node A has been replayed.
Threats: T-08, T-09, T-10
Challenges: C-05
Security solution:
a) The following solution may be used to provide replay protection for any content received by SOAP node B. Transport Layer Integrity (Section 5.1.1)
b) Currently there is no application interoperability solution at the SOAP message layer.
A SOAP message is sent over a SOAP message path from a SOAP node N0 through zero or more SOAP Intermediaries to a SOAP node Nk using a series of HTTP Requests.
This scenario applies to situations where the loss of individual SOAP messages is insignificant (for example, in a status monitoring scenario where periodic status update events are provided such that if one update event is lost, a subsequent update event will convey correct status). No SOAP message response is generated by Nk or expected by N0. Regardless of the protocol implemented by the transport layer, N0 receives no SOAP message response.
The transport layer may not guarantee delivery of the SOAP message. The N0 or any SOAP Intermediary may not be aware whether a SOAP message was successfully sent or delivered to, received or processed by, any other node. Receipt of an HTTP Response indicates that at the very least that the HTTP Node associated with the receiver has received the HTTP Request but does not guarantee that the SOAP message will ever arrive at the receiver.
SOAP Nodes:
HTTP Sessions:
SOAP Messages:
· M: N0 ® ... ® Nk
Hops:
· (for r = 1, ... k –1) Hr: Nr ® N1 (Session Sr )
Security Requirements
None beyond generic requirements of Section 6.4
This scenario is derived from the Synchronous Request/Response scenario in the WS-I Basic Applications Usage Scenarios [BPSA UsageScenarios]
A SOAP message (called the request) is sent from a SOAP node N0 through zero or more SOAP Intermediaries to a SOAP node Nk. A SOAP message called the response is sent by Nk to N0. The SOAP Path of this SOAP message is the reverse of that of the request. The Hops used in the transmission of the response are the HTTP responses to the Hops used in the transmission of the request.
SOAP Nodes:
Sessions:
SOAP Messages:
Hops:
Security Requirements
None beyond generic requirements of Section 6.4
This scenario was derived from the Basic Callback scenario in the WS-I Basic Sample Applications Usage Scenarios. [BPSA UsageScenarios]
The first SOAP Message APPLICATION-REQUEST is sent from Node A through zero or more to Node B through a series of Hops. APPLICATION-REQUEST contains information that indicates where B should send the APPLICATION-RESPONSE.
B sends a SOAP Message (acknowledgement) to A through the Http responses of the same set of Hops
After APPLICATION REQUEST is processed B sends a SOAP Message APPLICATION-RESPONSE to A through zero or more intermediaries through a series of Hops.
A sends a SOAP Message(acknowledgement) to B through the Http responses of the same set of Hops.
The APPLICATION-REQUEST and APPLICATION RESPONSE are related via correlation information that is provided by A in APPLICATION-REQUEST and duplicated by B into APPLICATION-RESPONSE.
SOAP Nodes:
Sessions:
SOAP Messages:
Hops:
Security Requirements:
Requirement: Message Correlation
SOAP Node A must be able to securely determine whether content of hop AP-RESPr+1 supplied by SOAP Node B was generated in response to APPLICATION-REQUEST. This requirement addresses the fact that related messages may be delivered on unrelated sessions.
Threats: T-01, T-03, T-04, T-05, T-06, T-09, T-10
Challenges: C-01, C-02, C-03, C-04
Security solutions:
Providing a solution for this requirement would require composition of a solution using techniques that are not described in the documents that are in scope for this profile.
An example of a solution would be for SOAP Node A to provide (with confidentiality, integrity and authentication) some correlation information X along with the content C. SOAP Node B would provide (with confidentiality, integrity and authentication) the same correlation information X along with the application level response.
Requirement: Node Correlation
SOAP Node A must be able to securely determine whether the content of AP-RESPr+1 was supplied by SOAP Node B in response to content C sent to SOAP Node B.
This requirement addresses the possibility that the credential Q used by SOAP Node A to identify SOAP Node B when targeting content to SOAP Node B is not the same credential R used by SOAP Node B to identify itself when targeting content to SOAP Node A.
Threats: T-01, T-03, T-04, T-05, T-06, T-09, T-10
Challenges: C-01, C-02, C-03, C-04
Security solution:
Providing a solution for this requirement would require composition of a solution using techniques that are not described in the documents that are in scope for this profile.
The simplest example of a solution, based on the example given for Message Correlation, would be to ensure that the same credential was used to provide confidentiality to, and authentication from, SOAP Node B (Q = R). A more complex solution, still based on the Message Correlation example, would require SOAP Node A to have access to some mapping of several credentials to SOAP Node B (Q => B and R => B).
This section contains discussions of security aspects that are not considered in the security requirements of the scenarios. It is included so that the reader is aware that these have not been overlooked. The primary reasons that they are not considered is that mechanisms to deal with them are not present within the technologies in the charter of this committee or because in some cases (e.g. Credentials Issuance) the solutions are not technological.
Definition: Non-repudiation: A security service that provides protection against false denial of involvement in a communication.
Explanation: Protection against false denial of an action associated with a Web service message. Non-repudiation technologies do not prevent repudiation, but rather provide evidence that may be used by a third party to resolve disputes.
Threat association: Accountability related threats along with threats associated with C-01, C-02 and C-03 must be addressed relative to this challenge and needs to be discussed further.
Definition: Credential(s): Data that is transferred or presented to establish either a claimed identity or the authorizations of a system entity.
Explanation: The process of initially providing a principal with a means of identifying itself, via online or offline mechanisms. Traditionally, “issuance” refers only to certificates, but here it is used for any information furnished by an authority that is willing to vouch for the principal. We believe that this security challenge is out of scope.
Creation of a credential via transformation from an existing credential to an equivalent one in another format is not issuance in the sense of this section.
Threat association: Out of scope
Note that out of scope threats are designated as T(OOS)-XX.
|
ID |
Name |
Description |
|
T(OOS)-01 |
Key Attack / Weak Algorithm |
The algorithm chosen is subject to attacks and/or the key(s) can be compromised. This covers a variety of attacks. Most of these have to do with details of the implementation or operational procedures, which is the reason for considering them to be outside the scope of a specification profile. However some aspects of profiles, e.g. selection of cryptographic algorithms, would be relevant to this threat. Here as elsewhere there are two levels: some parameter settings would be universally considered insecure, e.g. null encryption algorithm. In other cases, the choice would be a matter of local policy. For example, some organizations consider a 1024 bit RSA key adequately strong and others do not. Still others consider it satisfactory for some uses and not others. |
|
T(OOS)-02 |
Traffic Analysis |
By analyzing aspects of the messages such as its source, destination, size, frequency, etc., determinations can be made about potential contents (e.g. it is determined that one company may be trying to buy another). This has many subtle forms. For example, during WW II, Russian scientists deduced that the Americans were building an Atomic Bomb, because the physicists in question had stopped publishing papers. |
|
T(OOS)-03 |
Host Penetration/ Access |
Information is obtained by compromising a computer system (e.g. unauthorized access to a computer). Any threat analysis must assume some part of the system is secure. This is called the Trusted Computing Base (TCB). If there is no TCB, it is not possible to conclude anything about the behavior of the system, since presumably an attacker could modify its behavior at will. Thus, in a sense, this threat is out of scope of ANY design or specification, although certainly not out of scope of implementation and operations. |
|
T(OOS)-04 |
Network Penetration/ Access |
Information is obtained by compromising a computer network (e.g. unauthorized access to an internal network). This threat presumes a topological approach to security, e.g. firewalls or security gateways. If appropriately strong mechanisms are used on an end-to-end basis, network attacks are reduced to denial-of-service. Thus this threat is out of scope because it is essentially equivalent to the standard assumption of an untrusted network. |
|
T(OOS)-05 |
Timing |
By analyzing the time it takes to perform an action, information can be deduced (e.g. validity of a username, or key information). This is out of scope because it is an implementation issue rather than a specification issue. However, it should be noted that some published cryptographic timing attacks require timing measurements which are much smaller that the average variability of latency in typical networks and thus not of practical concern. |
|
T(OOS)-06 |
Covert Channels |
Information is conveyed outside of a secure perimeter by means of secret communication paths (e.g. by toggling an externally visible flag, secret information is conveyed). This threat is usually only consider seriously in military or intelligence environments. Typically the engineering approach taken is not to eliminate the channel, but to reduce its bandwidth to the point of being useless. |
|
T(OOS)-07 |
Message Archives |
By penetrating the queue of a store-and-forward SOAP intermediary, or the store of an archival system, information about a message can be discovered (e.g. a message in a store and forward queue can be discovered which otherwise wouldn't have been seen). Note that in many circumstances this is a variation on T(OOS)-03. The main reason for calling out this threat separately is because end-to-end message protection measures can counter it, whereas hop-by-hop measures cannot. |
|
T(OOS)-08 |
Network Spoofing |
A message is sent which appears to be from another machine (e.g. BadGuy sends a message which appears as though it is from GoodGuy). Comments similar to those under T(OOS)-04 apply here. If the message does not reach the application, there is little a profile of a specification can have to say about it. If it does reach the application, it is essentially the same as T-04 and T-06. |
|
T(OOS)-08 |
Trojan Horse |
Information is secretly passed along with the message that plants a Trojan horse (e.g. a message is added which is detected by planted software which causes special behaviors to occur). Note that this is a variation on T-01 and T-02. |
|
T(OOS)-09 |
Virus |
Information is secretly passed along with the message that plants a virus (e.g. a message is added which is detected by planted software which causes special behaviors to occur). Note that this is a variation on T-26. Viruses are usually planted by action of unsuspecting user or occasionally program flaw that triggers execution without user action. This can be contrasted with a Worm, which spreads itself autonomously without user action. Worms typically execute other threats found in this table in automated fashion. Some authorities have abandoned the distinction among various programmatic threats and use the term “malware” to cover all types. |
|
T(OOS)-10 |
Tunneling |
Information is secretly passed along with the message (e.g. a message is added which is detected by planted software which causes special behaviors to occur). Note that this is a variation on T-01 and T-02. |
|
T(OOS)-11 |
Denial of Service |
Silver Bullet: specific messages or command sequences causes failure. Almost invariably a result of implementation error, not design error. (Note that this can also result in a system or application compromise instead of merely a Denial of Service.) Inconceivable that a Profile would require dealing with this threat.
|
|
T(OOS)-12 |
Denial of Service |
Flooding: Sheer volume of message traffic overloads some critical resource, typically server or network link bandwidth. This is usually a configuration issue not a design issue. If the bogus traffic is truly indistinguishable from legitimate traffic there may be no defense. It is important to try to
|
|
T(OOS)-13 |
Repudiation |
A message is sent and then the sender denies having sent it. Achieving non-repudiation requires both technical and business aspects since a party may always claim a disconnect with the technology ("the software did it, not me, I didn't know").Public Key cryptographic systems have a special property that cannot be achieved by secret key systems without the use of a trusted third party. The property is that it is possible for a party to be able to verify something e.g. a digital signature, without being able to produce it themselves. When this technical property was first observed, it was called ”non-repudiation”. Much later it became widely believed that non-repudiation was a well-established legal concept (It is not.) and very desirable for electronic commerce. The confusion between the technical and legal meanings of this term continues. |
Table 4: Out of Scope Threats
HTTP – Hypertext Transfer Protocol
HTTPS – Hypertext Transfer Protocol Secure
IETF – Internet Engineering Task Force
MD5 – one Message-Digest algorithm (RFC-1321)
MEP – Message Exchange Pattern
MIME – Multipurpose Internet Mail Extensions
OASIS – not an acronym
OOS – Out Of Scope
RFC – Request for Comment (Used by IETF)
SCM – Supply Chain Management; the WS-I Sample Application for 1.0
SHA – Secure Hash Algorithm
SOAP - Simple Object Access Protocol
SSL – Secure Sockets Layer
TLS – Transport Layer Security
WS-Security – OASIS SOAP Message Security specifications
XML – Extensible Markup Language
X.509 – An ITU (International Telecommunication Union) standard for “certificates” Also known as ISO/IEC 9594-8:1988
[1] The tag naming convention consists of three parts. The first character is a “B” in the first character to identify that this is a binding level solution. (Note: “T” was not used because of possible confusion with “T” used by Threat tags.) The next 1 to 3 letters identify the transport challenge: “I” for Integrity, “S” for confidentiality (Secret), “P” for Provider authentication, and “C” for Consumer authentication. The last component is a number identifying the solution instance.
[2] Note: user can support NULL-SHA or NULL-MD5 cipher suites for this usage.
[3] The tag naming convention consists of three parts. The first character is a “S” in the first character to identify that this is a SOAP message level solution. The next letter identify the type of SOAP message level challenge: “I” for Integrity, “C” for Confidentiality, “A” for SOAP sender Authentication. The last component is a number identifying the solution instance.
[4] Kerberos tokens are part of our charter candidate technologies. However, usage of this technology in this profile will be deferred until OASIS TC deliver this core specification. Note: as other types of security tokens, e.g., SAML assertions or XrML tokens, are added to our list of charter technologies, they will be added to these security profiles.