SACM Working Group H. Birkholz
Internet-Draft Fraunhofer SIT
Intended status: Standards Track N. Cam-Winget
Expires: October 10, 2016 Cisco Systems
April 8, 2016

SACM Information Model
draft-camwinget-sacm-information-model-00

Abstract

***replaces abstract in WG IM*** This document defines the information model for Security Automation and Continuous Monitoring (SACM). This includes the definition of information elements transported between SACM components (data in motion) and how to express their relationships. This information model is maintained as the IANA “SACM Information Elements” registry. The information model captures the information needs described in the use cases defined by SACM.

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.

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."

This Internet-Draft will expire on October 10, 2016.

Copyright Notice

Copyright (c) 2016 IETF Trust and the persons identified as the document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.


Table of Contents

1. Introduction

***replaces Introduction in the WG IM*** The purpose of the SACM Information Model (IM) is to ensure interoperability between SACM data models that are used as transport encoding and to provide a base set of information elements that may be exposed or shared between SACM components in a scalable and extensible fashion. A complete set of requirements imposed on the IM can be found in [I-D.ietf-sacm-requirements]. The SACM IM defines information elements that are required to carry out the tasks conducted by SACM components. The SACM IM itself is intended to be used for data exchange between SACM components (data in motion). Nevertheless, the information elements defined in this document can be leveraged to create and align corresponding data models for data at rest.

The information model expresses, for example, target endpoint (TE) attributes, guidance or evaluation results. The corresponding information elements (IE) are consumed and produced by SACM components as they carry out tasks.

The primary tasks that this information model supports (on data, control and management plane) are:

2. Requirements notation

The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “NOT RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in RFC 2119, BCP 14 [RFC2119].

3. Information Elements (IE)

**to be inserted between section 2 and section 3** Every type or group of information, e.g. the information elements, defined in this document represent subjects transported (data in motion) between SACM components and are associated with a unique label in the information model: their name. This document defines a set of information elements standardized by SACM.

3.1. Context of Information Elements

The IE in this information model represent information related to the following areas (based on the use cases described in [RFC7632]):

3.2. Extensibility of Information Elements

A SACM data model based on this information model MAY include additional information elements that are not defined here. The labels of additional information elements included in different SACM data models MUST NOT conflict with the labels of the information elements defined by this information model, and the names of additional information elements MUST NOT conflict with each other or across multiple data models. In order to avoid naming conflicts, the labels of additional IEs SHOULD be prefixed to avoid collision across extensions. The prefix MUST include an organizational identifier and therefore, for example, MAY be an IANA enterprise number, a (partial) name space URI or an organization name abbreviation.

4. Structure of Information Elements

**replaces beginning text of Information Model Framework and 3.1-3.4, will move syntax 3.1.1 and 3.2.1 to aggregated sub-section, will also privacy sub-section 3.5 and label sub-section 3.6** The IEs defined in this document are differentiated into two basic types of Information Elements:

Metadata is constructed as a subject and is associated with attributes or subjects to provide additional information about them. The IM explicitly defines two specific kinds of metadata: metadata about the data origin and metadata about the data source. Metadata can include relationships that refer to other attributes or subjects by referencing labels included in their corresponding metadata.

4.1. Atomic Information Elements (AIE)

**to be salvaged an then removed** Atomic IEs represent the smallest building blocks for SACM content, including, for example, a SACM endpoint attribute, a policy entry, a configuration item, an expected states, or a threshold value. AIE can be bundled into composite IE. The set of AIEs defined by the SACM IM is described in section Section 5.2.

In essence, AIEs are attribute value pairs that constitute the “leaves” in a SACM semantic structure. While the SACM IM sometimes does elaborate on the structure of values (e.g. an IPv6 address is an octet string with a maximum length of 16 that my be collapsed in certain conditions), it does not prescribe specific types used in the data model representation (e.g. an unbounded character string).

Every AIE is registered as an corresponding entry at the IANA registry. The Integer Index of the IANA SMI number tables can be used by SACM data models.

4.2. Composite Information Elements (CIE)

**to be salvaged an then removed** Composite IEs constitute bundles of atomic AIEs and/or composite IEs. A CIE represents a specific set of related information that share a semantic relationship, e.g. a SACM statement metadata or state information about a network interface. The set of CIEs defined by the SACM IM is described in section Section 5.3. In essence, CIEs are a “named container” construct that can be used to compose additional CIEs that go beyond the ones standardized by the SACM information model.

The SACM IM allows for recursive or circular nesting of composite IEs. A SACM data Model (DM) MUST include the “default-depth” base AIE that is part of the SACM content metadata.

4.3. SACM Statements

**to be salvaged an then removed** The data exchanged between SACM components is always embedded in a SACM statement. SACM Statements contain one or more CIEs and/or AIEs. A SACM statement functions as an “envelope” type that is associated with metadata about the providing SACM component. The SACM statement metadata can be used to resolve conflicting information, retrace the provenance of information or to locate archived information in data repositories.

Examples of SACM statement metadata information elements:

SACM statements are comprised of one or more CIEs; Section 6 provides examples for constructing SACM statements.

4.4. SACM Content Elements

**to be salvaged an then removed** SACM Content Elements are categorized CIEs. The content elements can be composed of one or more AIEs and/or CIEs or it can be another representation that is embedded in the statement, for example, an IPFIX Template Record. Each SACM content element has its own Content Metadata associated with it (analogously to the way that each SACM statement has metadata associated to it). Content element metadata include information about its type, data source (the result produced by a collector) or data origin (the result produced by most other SACM components).

Examples of SACM content element metadata information elements:

SACM content elements are described in section FIXME.

4.5. Relationship Types

**to be salvaged an then removed** Relationships are expressed via AIE contained within a CIE. There are two ways SACM content elements are associated with each other. “A Flow” associated with “A User”, for example, would be a typical case, in which two separate SACM content elements could be associated with each other.

One way is to include the Relationships AIE in the content element metadata that preludes the actual content (in this example, the content element metadata of the flow record). Relationship Types are uni-directional. For example, the “is-associated-with-user” Relationship AIE included in the content element metadata points to a specific user via a corresponding content element identifier.

The alternative way is to include the reference of associated information directly into the content of the content element. A session CIE, for instance, could refer to a specific user by including identifying attributes about that user. While this is a valid way of creating a relationship between different kinds of content, it requires careful matching or the introduction of another appropriate identifier mechanism (that does not conflict with other SACM statements and SACM content element identifiers). If a SACM data model allows for transport of other representations as payload of a content element (e.g. a pcap fragment containing suspicious packets, for example), there might be no alternative as to use the content element metadata to include relationships to other content elements.

4.6. Events

**to be salvaged an then removed** Events are a specific type of CIE that are always associated with a time stamp and represent a change of state or configuration that can be expressed as a SACM content. The time an event was published by a SACM component is recorded in its corresponding SACM statement metadata, the time it was created (or initially observed) is recorded in its content element metadata. It is also recorded in the CIE itself, which is somewhat redundant but can improve performance in some scenarios. Event CIE can also include the past state or configuration before the change occurred, or – if applicable – a threshold or trigger condition that lead to the creation of the event.

5. Information Element Vocabulary

**to be inserted in section 5 as candidates** The vocabulary of Information Element names standardized by the SACM IM does not prescribe the use of these exact same names in every SACM data model. If terms diverge, a mapping has to be provided in the corresponding SACM data model document.

A subset of the names of the information elements defined in this document are appended with “-type”. This indicates that the IM defines a set of values for these information elements (e.g. the interface types defined by the IANA registry or the relationship types).

5.1. Vocabulary of Categories

Categories are special Information Elements that enable to refer to multiple types of IEs via just one name. Therefore, they are similar to a type-choice. A prominent example of a category is network-address. Network-address is a category that every kind of network address is associated with, e.g. mac-address, ipv4-address, ipv6-address, or typed-network-address. If a CIE includes network-address as one of its components, any of that categories members is valid to be used in its stead.

Another prominent example is EndpointIdentifier. Some IEs can be used to identify (and over time re-recognize) target endpoints - those are associated with the category endpoint-identifier.

content:
this is a very broad category. Content is the payload of a content element in a SACM statement. Formally, metadata is the complement to content and everything that is not part of SACM statement metadata or content element metadata is therefore considered to be content. Every IE can be content (although the same type of IE can be used in the metadata at the same time - and those would not be content as described before). Annotating every IE with this category would be highly redundant and is therefore omitted for brevity.
network-address: (work-in-progress)
ipv4-address
ipv6-address
mac-address

endpoint-identifier: (work-in-progress)

software-component: (work-in-progress)

software-label: (work-in-progress)

5.2. Vocabulary of Atomic Information Elements

**to be inserted in section 5 as candidates** The content of every Atomic Information Element is expressed in a single value. Note that while this section lists AIEs, some of them may also be represented as a CIE (especially if metadata is used).

access-privilege-type:
a set of types that represents access privileges (e.g. read, write, none)
References: none
account-name:
a label that uniquely identifies an account that can require some form of (user) authentication to access
References: none
administrative-domain:
a label the is supposed to uniquely identify an administrative domain
References [IFMAP]
address-association-type:
a set of types that defines the type of address associations (e.g. broadcast-domain-member-list, ip-subnet-member-list, ip-mac, shared-backhaul-interface, etc.)
References: none
address-mask-value:
a value that expresses a generic address subnetting bitmask
address-type:
a set of types that specifies the type of address that is expressed in an address CIE (e.g. ethernet, modbus, zigbee)
References: none
address-value:
a value that expresses a generic network address
References: none
Category: network-address
application-component:
a label that references a “sub”-application that is part of the application (e.g. an add-on, a cipher-suite, a library)
References: [SWID]
Category: software-component
application-label:
a label that is supposed to uniquely reference an application
References: [SWID]
Category: software-label
application-type:
a set of types (FIXME maybe a finite set is not realistic here - value not enumerator?) that identifies the type of (user-space) application (e.g. text-editor, policy-editor, service-client, service-server, calendar, rouge-like RPG)
References: [SWID]
Category: software-type
application-manufacturer:
the name of the vendor that created the application
References: [SWID]
Category: software-manufacturer
application-name:
a value that represents the name of an application given by the manufacturer
References: [SWID]
application-version:
a version string that identifies a specific version of an application
References: [SWID]
Category: software-version
authenticator:
a label that references a SACM component that can authenticate target endpoints (can be used in a target-endpoint CIE to express that the target endpoint was authenticated by that SACM component)
References: none
attribute-name:
a value that can express the attribute name of generic Attribute-Value-Pair CIE
References: none
attribute-value:
a value that can express the attribute value of generic Attribute-Value-Pair CIE
References: none
authentication-type:
a set of types that expresses which type of authentication was used to enable a network interaction/connection
References: [PXGRID]
birthdate:
a label for the registered day of birth of a natural person (e.g. the date of birth of a person as an ISO date string http://rs.tdwg.org/ontology/voc/Person#birthdate)
References: [SCAP-AI]
bytes-received:
a value that represents a number of octets received on a network interface
Reference : [PXGRID]
bytes-sent:
a value that represents a number of octets sent on a network interface
Reference : [PXGRID]
certificate:
a value that expresses a certificate that can be collected from a target endpoint
References: none
Category: endpoint-identifier
collection-task-type:
a set of types that defines how collected SACM content was acquired (e.g. network-observation, remote-acquisition, self-reported)
Reference: none
confidence:
a representation of the subjective probability that the assessed value is correct. If no confidence value is given it is assumed that the confidence is 1 (limits confidence values to the range between zero and one)
References: [ARF]
content-action:
a set of types that expresses a type of action (e.g. add, delete, update). Can be associated, for instance, with an event CIE or with an network observation
References: [ARF]
content-elements:
a value that represents the number of content-elements included in a SACM statement
References: none
content-topic:
a set of types that defines what kind of concept the information is included in a content element (e.g. Session, User, Interface, PostureProfile, Flow, PostureAssessment, TargetEndpoint)
References: none
content-type:
a set of types that defines what kind of information is included in a content element (e.g. EndpointConfiguration, EndpointState, DirectoryEntry, Event, Incident)
References: none
country-code:
a set of types according to ISO 3166-1 trigraphic codes of countries
References: FIXME
data-origin:
a label that uniquely identifies a SACM component in and across SACM domains
References: none
Aliases: sacm-component-id
data-source:
a label that is supposed to uniquely identify the data source (e.g. a target endpoint or sensor) that provided an initial endpoint attribute record
References: [ARF]
Aliases: te-id (work-in-progress)
decimal-fraction-denominator:
a denominator value to express a decimal fraction time stamp (e.g. in timestamp)
References: none
decimal-fraction-numerator:
a numerator value to express a decimal fraction time stamp (e.g. in timestamp)
default-depth:
a value that expresses how often a circular reference of CIE is allowed to repeat, or how deep a recursive nesting may occur, respectively.
References: none
discoverer:
a label that refers to the SACM component that discovered a target endpoint (can be used in a target-endpoint CIE to express, for example, that the target endpoint was authenticated by that SACM component)
References: none
email-address:
a value that expresses an email-address
References: none
event-type:
a set of types that define the categories of an event (e.g. access-level-change, change-of-privilege, change-of-authorization, environmental-event, or provisioning-event)
Reference: none
event-threshold:
if applicable, a value that can be included in an event CIE to indicate what numeric threshold value was crossed to trigger that event
Reference: none
event-threshold-name:
if an event is created due to a crossed threshold, the threshold might have a name associated with it that can be expressed via this value
References: none
event-trigger:
this value is used to express more complex trigger conditions that may cause the creation of an event.
firmware-id:
a label that represents the BIOS or firmware ID of a specific target endpoint
Reference: none
Category: endpoint-identifier
hardware-serial-number:
a value that identifies a piece of hardware that is a component of a composite target endpoint (in essence, every target endpoint is a composite) and can be acquired from a target endpoint by a collection task
Reference: none
Category: endpoint-identifier
host-name:
a label typically associated with an endpoint but not always intended to be unique in a given scope
References [ARF], [SCAP-AI]
Category: endpoint-identifier
interface-label:
a unique label a network interface can be referenced with
Reference: none
ipv6-address-subnet-mask-cidrnot:
an IPv6 subnet bit mask in CIDR notation
References: TBD
ipv6-address-value:
an IPv4 address value
References: TBD
Category: endpoint-identifier, network-address
ipv4-address-subnet-mask-cidrnot:
an IPv4 subnet bit mask in CIDR notation
References: TBD
ipv4-address-subnet-mask:
an IPv4 subnet mask
References: TBD
ipv4-address-value:
an IPv4 address value
References: TBD
Category: endpoint-identifier, network-address
layer2-interface-type:
a set of types referenced by IANA ifType
References: [RFC3635], [RFC2863]
layer4-port-address:
a layer 4 port address (typically used, for example, with TCP and UDP)
References: none
Category: network-address
layer4-protocol:
a set of types that express a layer 4 protocol (e.g. UDP or TCP)
location-name:
a value that represents a named region of space FIXME
References: [IFMAP], [ARF], [SCAP-AI]
mac-address:
a value that expresses an Ethernet address
References: [IFMAP], [ARF], [SCAP-AI]
Category: endpoint-identifier, network-address
method-label:
a label that references a specific method registered and used in a SACM domain (e.g. method to match and re-identify target endpoints via identifying attributes)
References: none
method-repository:
a label that references a SACM component methods can be registered at and that can provide guidance in the form of registered methods to other SACM components
References: none
network-access-level-type:
a set of types that expresses categories of network access-levels (e.g. block, quarantine, etc.)
References: [IFMAP]
network-id:
most networks, such as AS, an OSBF domains, or vlans, can have an ID that is represented via this AIE
References: none
network-interface-name:
a label that uniquely identifies an interface associated with a distinguishable endpoint
References: FIXME
network-layer:
a set of layers that express the specific network layer an interface operate on (typically layer 2-4)
References: FIXME
network-name:
a label that is associated with a network. Some networks, for example effective layer2-broadcast-domains, are difficult to “grasp” and therefore quite complicated to name
References: none
organization-id:
a label that is supposed to uniquely identify an organization
References: [ARF]
organization-name:
a value that represents the name of an organization
References: [ARF]
os-component:
a label that references a “sub-component” that is part of the operating system (e.g. a kernel module, microcode, or ACPI table)
References: [SWID]
Category: software-component
os-label:
a label that references a specific version of an operating system, including patches and hotfixes
References: [SWID]
Category: software-label
os-manufacturer:
the name of the manufacturer of an operating system
References: [IFMAP]
Category: software-manufacturer
os-name:
the name of an operating system
References: [IFMAP]
Category: software-name
os-type:
a set of types that identifies the type of an operating system (e.g. real-time, security-enhanced, consumer, server)
References: none
Category: software-type
os-version:
a value that represents the version of an operating-system
Category: software-version
patch-id:
a label the uniquely identifies a specific software patch
References: [ARF]
patch-name:
the vendor’s name of a software patch
References: [ARF], [SWID]
person-first-name:
the first name of a natural person
References: [ARF], [SCAP-AI]
person-last-name:
the last name of a natural person
References: [ARF], [SCAP-AI]
person-middle-name:
the first name of a natural person
References: [ARF], [SCAP-AI]
phone-number:
a label that expresses the u.s. national phone number (e.g. pattern value=”((\d{3}) )?\d{3}-\d{4}”)
References: [ARF], [SCAP-AI]
phone-number-type:
a set of types that express the type of a phone number (e.g. DSN, Fax, Home, Mobile, Pager, Secure, Unsecure, Work, Other)
References: [ARF]
privilege-name:
the attribute-name of the privilege represented as an AVP
References: none
privilege-value:
the value-content of the privilege represented as an AVP
References: none
protocol:
a set of types that defines specific protocols above layer 4 (e.g. http, https, dns, ipp, or unknown)
References: none
public-key:
the value of a public key (regardless of its method of creation, crypto-system, or signature scheme) that can be collected from a target endpoint
Reference: none
Category: endpoint-identifier
relationship-content-element-guid:
a reference to a specific content element used in a relationship CIE
References: none
relationship-statement-guid:
a reference to a specific SACM statement used in a relationship CIE
References: none
relationship-object-label:
a reference to a specific label used in content (e.g. a te-label or a user-id). This reference is typically used if matching content AIE can be done efficiently and can also be included in addition to a relationship-content-element-guid reference.
References: none
relationship-type:
a set of types that is in every instance of a relationship CIE to highlight what kind of relationship exists between the CIE the relationship is included in (e.g. associated_with_user, applies_to_session, seen_on_interface, associated_with_flow, contains_virtual_device)
References: none
role-name:
a label that references a collection of privileges assigned to a specific entity (identity? FIXME)
References: FIXME
session-state-type:
a set of types a discernible session (an ongoing network interaction) can be in (e.g. Authenticating, Authenticated, Postured, Started, Disconnected)
References: [PXGRID]
statement-guid:
a label that expresses a global unique ID referencing a specific SACM statement that was produced by a SACM component
References: none
statement-type:
a set of types that define the type of content that is included in a SACM statement (e.g. Observation, DirectoryContent, Correlation, Assessment, Guidance)
References: none
status:
a set of types that defines possible result values for a finding in general (e.g. true, false, error, unknown, not applicable, not evaluated)
References: [ARF]
sub-administrative-domain:
a label for related child domains an administrative domain can be composed of (used in the CIE administrative-domain)
References: none
sub-interface-label:
a unique label a sub network interface (e.g. a tagged vlan on a trunk) can be referenced with
References: none
super-administrative-domain:
a label for related parent domains an administrative domain is part of (used in the CIE administrative-domain)
References: none
super-interface-label:
a unique label a super network interface (e.g. a physical interface a tunnel interface terminates on) can be referenced with
References: none
te-assessment-state:
a set of types that defines the state of assessment of a target-endpoint (e.g. in-discovery, discovered, in-classification, classified, in-assessment, assessed)
References: [ARF]
te-label:
an identifying label created from a set of identifying attributes used to reference a specific target endpoint
References: none
te-id:
an identifying label that is created randomly, is supposed to be unique, and used to reference a specific target endpoint
References: [ARF], [SWID]
Aliases: data-source
timestamp:
a timestamp the expresses a specific point in time
References: [IFMAP], [ARF]
timestamp-type:
a set of types that express what type of action or event happened at that point of time (e.g. discovered, classified, collected, published). Can be included in a generic timestamp CIE
References: none
units-received:
a value that represents a number of units (e.g. frames, packets, cells or segments) received on a network interface
Reference : [PXGRID]
units-sent:
a value that represents a number of units (e.g. frames, packets, cells or segments) sent on a network interface
Reference : [PXGRID]
username:
a part of the credentials required to access an account that can be collected from a target endpoint
References: none
Category: endpoint-identifier
user-directory:
a label that identifies a specific type of user-directory (e.g. ldap, active-directory, local-user)
Reference: [PXGRID]
user-id:
a label that references a specific user known in a SACM domain
References: [PXGRID]
web-site:
a URI that references a web-site
References: [ARF]
WGS84-longitude:
a label that represents WGS 84 rev 2004 longitude
References: [SCAP-AI]
WGS84-latitude:
a label that represents WGS 84 rev 2004 latitude
References: [SCAP-AI]
WGS84-altitude:
a label that represents WGS 84 rev 2004 altitude
References: [SCAP-AI]

5.3. Vocabulary of Composite Information Elements

**to be inserted in section 5 as candidates** The content of every Composite Information Element is expressed by the mandatory and optional IE it can be composed of. The components of an CIE can have a cardinality associated with them:

If there is no cardinality highlighted or the cardinality (+) or (n*m) is used, including this IE in the CIE is mandatory. In contrast, optional IE are expressed via the cardinality (?) or (*). An CIE can prescribe a strict sequence to the component IE it contains. This in indicated by an (s).

address-association (s):
some addresses are associated with each other, e.g. a mac-address can be associated with a number of IP addresses or a sensor address can be associated with the external address of its two redundant IP gateways. The first address is the address a number of addresses with the same type is associated with. An address type SHOULD be included and the addresses associated with the first address entry MUST be of the same type. NANCY FIXME
address
address-type (?)
address (+)
address-type (?)
administrative-domain:
this CIE is intended to express more complex setups of interconnected administrative domains
administrative-domain
sub-administrative-domain (*)
super-administrative-domain (?)
location (?)
application:
an application is software that is not part of the kernel space (therefore typically runs in the user space. An application can depend on specific running party of an operating system.
application-label (?)
application-name
application-type (*)
application-component (*)
application-manufacturer (?)
application-version (?)
application-instance:
a specific instance of an application that is installed on an endpoint. The application-label is used to refer to corresponding information stored in an application CIE
application-label
target-endpoint
attribute-value-pair:
a generic CIE that is used to express various AVP (e.g. Radius Attributes)
attribute-name
attribute-value
content-creation-timestamp:
a decimal fraction timestamp that specifies the point in time the content element was created by a SACM component
decimal-fraction-denominator
decimal-fraction-numerator
content-element:
content produced by a SACM component is encapsulated in content-elements that also include content-metadata regarding that content
content-metadata (+)
content (+)
content-metadata:
metadata regarding the content included in a specific content-element. The content the metadata annotates can be initially collected content - in this case a data-source has to be included in the metadata. Content can also be the product of a SACM component (e.g. an evaluator), which requires a data-origin IE instead that references the producer of information.
content-element-guid
content-creation-timestamp
content-topic
content-type
data-source (?)
data-origin (?)
relationship (*)
data-source:
a CIE that refers to a target endpoint that is the source of SACM content - either via a label (data-source, which could also be used without this CIE), or via a list of endpoint-identifiers (category). Both can be included at the same time but MUST NOT conflict.
data-source (?)
endpoint-identifier (*)
dst-flow-element:
identifies the destination of a flow. The port number SHOULD be included if the network-address is an IP-address.
network-address
layer4-port-address (?)
ethernet-interface:
the only two mandatory component of this CIE is the mac-address and the generated label (to distinguish non-unique addresses). This acknowledges the fact that in many cases this is the only information available about an Ethernet interface. If there is more detail information available it MUST be included to avoid ambiguity and to increase the usefulness for consumer of information. The exception are sub-interface-labels and super-interface-labels, which SHOULD be included.
interface-label
network-interface-name (?)
mac-address
network-name (?)
network-id (?)
layer2-interface-type (?)
sub-interface-label (*)
super-interface-label (*)
event (s):
this a special purpose CIE that represents the change of content. As with content-elements basically every content can be included in the two content entries. The mandatory content entry represents the “after” state of the content and the optional content entry can represent the “before” state if available or required.
event-type (?)
event-threshold (?)
event-threshold-name (?)
event-trigger (?)
typed-timestamp
content
content (?)
flow-record:
a composite that expresses a single flow and its statistics. If applicable, protocol and layer4-protocol SHOULD be included
src-flow-element
dst-flow-element
protocol (?)
layer4-protocol (?)
flow-statistics
flow-statistics:
this CIE aggregates bytes and units send and received
bytes-received
bytes-sent
units-received
units-sent
group:
insert text here (work in progress)
ipv4-address:
an IPv4 address is always associated with a subnet. This CIE combines these both tightly nit values. Either a subnet mask or a CIDR notation bitmask SHOULD be included.
ipv4-address-value
ipv4-address-subnet-mask-cidrnot (?)
ipv4-address-subnet-mask (?)
ipv6-address:
an IPv6 address is always associated with a subnet. This CIE combines these both tightly nit values. A CIDR notation bitmask SHOULD be included.
ipv6-address-value
ipv6-address-subnet-mask-cidrnot (?)
location:
a CIE that aggregates potential details about a location
location-name
WGS84-longitude
WGS84-latitude
WGS84-altitude
operation-system:
an operation-system is software that is directly interacting with the hardware, provides the runtime environment for the user-space and corresponding interfaces to hardware functions.
os-label (?)
os-name
os-type (*)
os-component (*)
os-manufacturer (?)
os-version (?)
organization:
this CIE aggregates information about an organization and can be references via its id
organization-id
organization-name
location (?)
person:
a CIE that aggregates the details about a person and combines it with a identifier unique to SACM domains
person-first-name
person-last-name
person-middle-name (*)
phone-contact (*)
email-address (*)
phone-contact:
this CIE can be used to reference a phone number and how it functions as a contact
phone-number
phone-number-type (?)
privilege:
a CIE to express privileges via a specific name/value pair
privilege-name
privilege-value
relationship:
the relationship CIE enables to associate the CIE it is included in with other CIE if they contain a unique identifier or label - providing an alternative to including attributes of other content CIE as a means to map them (which remains a valid alternative, though). The relationship CIE MUST at least reference one relationship object (either a SACM statement iden
relationship-type
relationship-content-element-guid (*)
relationship-statement-guid (*)
relationship-object-label (*)
sacm-statement:
every SACM components produces information in this format. This CIE can be considered the root IE for every SACM message generated. There MUST be at least one content element included in a SACM statement and if there are more than one, they are ordered in a sequence.
statement-metadata
content-element (+)(s)
session:
represents an ongoing network interaction that can be in various states of authentication or assessement
session-state-type
(work-in-progress)
src-flow-element:
identifies the source of a flow. The port number SHOULD be included if the network-address is an IP-address.
network-address
layer4-port-address (?)
statement-creation-timestamp:
a decimal fraction timestamp that specifies the point in time the SACM statement was created by a SACM component
decimal-fraction-denominator
decimal-fraction-numerator
statement-publish-timestamp:
a decimal fraction timestamp that specifies the point in time the SACM component attempted to publish the SACM statement (if successful, this will result in the publish-timestamp send with the SACM statement).
decimal-fraction-denominator
decimal-fraction-numerator
statement-metadata:
every SACM statement includes statement metadata about the SACM component it was produced by and a general category that indicates what this statement is about
statement-guid
data-origin
statement-creation-timestamp (?)
statement-publish-timestamp
statement-type
content-elements
target-endpoint:
this is a central CIE used in the process chains a SACM domain can compose. Theoretically every kind of information can be associated with a target endpoint CIE via its corresponding content element. A few select IE can be stored in the CIE itself to reduce the overhead of following references that would occur in most scenarios. If the hostname is unknown the value has to be set as an equivalent to “not available” (e.g. NULL). Comment from the authors: This is “work in progress” an a good basis for discussion
host-name
te-label
administrative-domain (?)
application-instance (*)
ethernet-interface (*)
address-association (*)
data-source (?)
operation-system (?)
te-profile:
a set of expected states, policies and pieces of guidance that can be matched to a target endpoint (or a class of target endpoints “work in progress”)
typed-timestamp:
a flexible timestamp CIE that can express the specific type of timestamp via its content. This is an alternative to the “named” timestamps that do not include a timestamp-type
decimal-fraction-denominator
decimal-fraction-numerator
timestamp-type
user:
a CIE that references details of a specific user known in a SACM domain active on a specific target endpoint
user-id
username (?)
data-source (?)
user-directory (?)

6. Example composition of SACM statements

This section illustrates how SACM statements can be composed of content information elements, how relationship CIEs can be used in content metadata, and how the categories statement-type, content-topic and content-type are intended to be used.

The SACM statements instances are written in pseudo code. AIE end with a colon. Some AIE include exemplary values to, for example, present how references to guid and labels can be used. For the sake of brevity, not all mandatory IE that are part of a CIE are always included (e.g. as it is the case with target-endpoint).

The example shows three SACM statements that were produced by three different SACM components that overall include four related content elements.

This is (work in progress).

sacm statement
  statement-metadata
    statement-guid: example-sguid-one
    data-origin: SACM-component-label-one
    statement-publish-timestamp: exmample-TS-one
    statement-type: Observation
  content-element
    content-metadata
      content-element-guid: example-cguid-one
      content-creation-timestamp:
      content-topic: Flow
      content-type: EndpointState
      relationship
        relationship-type: is-associated-with-user
        relationship-content-object: example-cguid-three
      relationship
        relationship-type: is-associated-with-te
        relationship-content-object: example-cguid-two
      relationship
        relationship-type: is-associated-with-te
        relationship-content-object: example-te-label      
    flow-record
      src-flow-element
        network-address (ipv4-address)
          ipv4-address-value:
          ipv4-address-subnet-mask-cidrnot:
        layer4-port-address: 23111
      dst-flow-element
        network-address (IPv4-address)
          ipv4-address-value:
          ipv4-address-subnet-mask-cidrnot:
        layer4-port-address: 22
      protocol: ssh
      layer4-protocol: tcp
      flow-statistics
        bytes-received:
        bytes-sent:
        units-received:
        units-sent:
  content-element
    content-metadata
      content-element-guid: example-cguid-two
      content-creation-timestamp:
      content-topic: TargetEndpoint
      content-type: EndpointConfiguration
    target-endpoint
      te-label: example-te-label
      host-name: example-host-name
      ethernet-interface: example-interface

sacm statement
  statement-metadata
    statement-guid: example-sguid-two
    data-origin: SACM-component-label-two
    statement-publish-timestamp: exmample-TS-two
    statement-type: DirectoryContent
  content-element
    content-metadata
      content-element-guid: example-cguid-three
      content-creation-timestamp:
      content-topic: User
      content-type: DirectoryEntry
    user
      user-name: example-username
      user-directory: component-id

sacm statement
  statement-metadata
    statement-guid: example-sguid-three
    data-origin: SACM-component-label-three
    statement-publish-timestamp: exmample-TS-three
    statement-type: Observation
  content-element
    content-metadata
      content-element-guid: example-cguid-four
      content-creation-timestamp:
      content-topic: Privileges
      content-type: Event
      relationship
        relationship-type: is-associated-with-user
        relationship-content-object: example-cguid-three 
    event
      event-type: change-of-privilege
      typed-timestamp
        decimal-fraction-denominator:
        decimal-fraction-numerator:
        timestamp-type: time-of-observation
      privilege
        privilege-name: super-user-escalation
        privilege-value: true
      privilege
        privilege-name: super-user-escalation
        privilege-value: false

7. IANA considerations

This document includes requests to IANA.

8. Security Considerations

9. Acknowledgements

10. Change Log

First revision -00

Second revision -00. Rename to Camwinget (removed -) to make submissions happier. Demonstrate how to integrate with WG draft.

11. Contributors

12. References

12.1. Normative References

, ", "
[ARF] Corporation., T., "Assessment Results Format", 2010.
[IFMAP]TCG Trusted Network Communications - TNC IF-MAP Metadata for Network Security Specification Version 1.1r9", May 2012.
[PXGRID] Appala, S., Cam-Winget, N., McGrew, D. and J. Verma, "An Actionable Threat Intelligence system using a Publish-Subscribe communications model", ACM Proceedings of the 2nd ACM Workshop on Information Sharing and Collaborative Security, page 61-70, DOI 10.1145/2808128.2808131, ISBN 978-1-4503-3822-6
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997.
[RFC2863] McCloghrie, K. and F. Kastenholz, The Interfaces Group MIB", RFC 2863, DOI 10.17487/RFC2863, June 2000.
[RFC3635] Flick, J., "Definitions of Managed Objects for the Ethernet-like Interface Types", RFC 3635, DOI 10.17487/RFC3635, September 2003.
[SCAP-AI] Wunder, J., Halbardier, A. and D. Waltermire, "Specification for Asset Identification 1.1", NIST Interagency Report 7693 , 2011.
[SWID]Information technology - Software asset management - Part 2: Software identification tag'", ISO/IEC 19770-2:2015, October 2015.

12.2. Informative References

[I-D.ietf-sacm-requirements] Cam-Winget, N. and L. Lorenzin, "Security Automation and Continuous Monitoring (SACM) Requirements", Internet-Draft draft-ietf-sacm-requirements-13, March 2016.
[RFC7632] Waltermire, D. and D. Harrington, "Endpoint Security Posture Assessment: Enterprise Use Cases", RFC 7632, DOI 10.17487/RFC7632, September 2015.

Authors' Addresses

Henk Birkholz Fraunhofer SIT Rheinstrasse 75 Darmstadt, 64295 Germany EMail: henk.birkholz@sit.fraunhofer.de
Nancy Cam-Winget Cisco Systems 3550 Cisco Way San Jose, CA 95134 USA EMail: ncamwing@cisco.com