This document is obsolete. Please
refer to RFC 7970.
Network Working Group R. Danyliw Request for Comments: 5070 CERT Category: Standards Track J. Meijer UNINETT Y. Demchenko University of Amsterdam December 2007
The Incident Object Description Exchange Format
Status of This Memo
This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited.
Abstract
The Incident Object Description Exchange Format (IODEF) defines a data representation that provides a framework for sharing information commonly exchanged by Computer Security Incident Response Teams (CSIRTs) about computer security incidents. This document describes the information model for the IODEF and provides an associated data model specified with XML Schema.
Organizations require help from other parties to mitigate malicious activity targeting their network and to gain insight into potential threats. This coordination might entail working with an ISP to filter attack traffic, contacting a remote site to take down a bot- network, or sharing watch-lists of known malicious IP addresses in a consortium.
The Incident Object Description Exchange Format (IODEF) is a format for representing computer security information commonly exchanged between Computer Security Incident Response Teams (CSIRTs). It provides an XML representation for conveying incident information across administrative domains between parties that have an operational responsibility of remediation or a watch-and-warning over a defined constituency. The data model encodes information about hosts, networks, and the services running on these systems; attack methodology and associated forensic evidence; impact of the activity; and limited approaches for documenting workflow.
The overriding purpose of the IODEF is to enhance the operational capabilities of CSIRTs. Community adoption of the IODEF provides an improved ability to resolve incidents and convey situational awareness by simplifying collaboration and data sharing. This structured format provided by the IODEF allows for:
o increased automation in processing of incident data, since the resources of security analysts to parse free-form textual documents will be reduced;
o decreased effort in normalizing similar data (even when highly structured) from different sources; and
o a common format on which to build interoperable tools for incident handling and subsequent analysis, specifically when data comes from multiple constituencies.
Coordinating with other CSIRTs is not strictly a technical problem. There are numerous procedural, trust, and legal considerations that might prevent an organization from sharing information. The IODEF does not attempt to address them. However, operational implementations of the IODEF will need to consider this broader context.
Sections 3 and 8 specify the IODEF data model with text and an XML schema. The types used by the data model are covered in Section 2. Processing considerations, the handling of extensions, and internationalization issues related to the data model are covered in
Danyliw, et al. Standards Track [Page 4]
RFC 5070 IODEF December 2007
Sections 4, 5, and 6, respectively. Examples are listed in Section 7. Section 1 provides the background for the IODEF, and Section 9 documents the security considerations.
The key words "MUST," "MUST NOT," "REQUIRED," "SHALL," "SHALL NOT," "SHOULD," "SHOULD NOT," "RECOMMENDED," "MAY," and "OPTIONAL" in this document are to be interpreted as described in RFC2119 [6].
Definitions for some of the common computer security-related terminology used in this document can be found in Section 2 of [16].
The normative IODEF data model is specified with the text in Section 3 and the XML schema in Section 8. To help in the understanding of the data elements, Section 3 also depicts the underlying information model using Unified Modeling Language (UML). This abstract presentation of the IODEF is not normative.
For clarity in this document, the term "XML document" will be used when referring generically to any instance of an XML document. The term "IODEF document" will be used to refer to specific elements and attributes of the IODEF schema. The terms "class" and "element" will be used interchangeably to reference either the corresponding data element in the information or data models, respectively.
The IODEF data model is a data representation that provides a framework for sharing information commonly exchanged by CSIRTs about computer security incidents. A number of considerations were made in the design of the data model.
o The data model serves as a transport format. Therefore, its specific representation is not the optimal representation for on- disk storage, long-term archiving, or in-memory processing.
o As there is no precise widely agreed upon definition for an incident, the data model does not attempt to dictate one through its implementation. Rather, a broad understanding is assumed in the IODEF that is flexible enough to encompass most operators.
o Describing an incident for all definitions would require an extremely complex data model. Therefore, the IODEF only intends to be a framework to convey commonly exchanged incident information. It ensures that there are ample mechanisms for
Danyliw, et al. Standards Track [Page 5]
RFC 5070 IODEF December 2007
extensibility to support organization-specific information, and techniques to reference information kept outside of the explicit data model.
o The domain of security analysis is not fully standardized and must rely on free-form textual descriptions. The IODEF attempts to strike a balance between supporting this free-form content, while still allowing automated processing of incident information.
o The IODEF is only one of several security relevant data representations being standardized. Attempts were made to ensure they were complimentary. The data model of the Intrusion Detection Message Exchange Format [17] influenced the design of the IODEF.
Further discussion of the desirable properties for the IODEF can be found in the Requirements for the Format for Incident Information Exchange (FINE) [16].
The IODEF implementation is specified as an Extensible Markup Language (XML) [1] Schema [2] in Section 8.
Implementing the IODEF in XML provides numerous advantages. Its extensibility makes it ideal for specifying a data encoding framework that supports various character encodings. Likewise, the abundance of related technologies (e.g., XSL, XPath, XML-Signature) makes for simplified manipulation. However, XML is fundamentally a text representation, which makes it inherently inefficient when binary data must be embedded or large volumes of data must be exchanged.
The various data elements of the IODEF data model are typed. This section discusses these data types. When possible, native Schema data types were adopted, but for more complicated formats, regular expressions (see Appendix F of [3]) or external standards were used.
A single character is represented by the CHARACTER data type. A character string is represented by the STRING data type. Special characters must be encoded using entity references. See Section 4.1.
The CHARACTER and STRING data types are implement as an "xs:string" [3] in the schema.
STRING data that represents multi-character attributes in a language different than the default encoding of the document is of the ML_STRING data type.
The ML_STRING data type is implemented as an "iodef:MLStringType" in the schema.
A binary octet is represented by the BYTE data type. A sequence of binary octets is represented by the BYTE[] data type. These octets are encoded using base64.
The BYTE data type is implemented as an "xs:base64Binary" [3] in the schema.
Enumerated types are represented by the ENUM data type, and consist of an ordered list of acceptable values. Each value has a representative keyword. Within the IODEF schema, the enumerated type keywords are used as attribute values.
The ENUM data type is implemented as a series of "xs:NMTOKEN" in the schema.
A timezone offset from UTC is represented by the TIMEZONE data type. It is formatted according to the following regular expression: "Z|[\+\-](0[0-9]|1[0-4]):[0-5][0-9]".
The TIMEZONE data type is implemented as an "xs:string" with a regular expression constraint in the schema. This regular expression is identical to the timezone representation implemented in an "xs: dateTime".
A list of network ports are represented by the PORTLIST data type. A PORTLIST consists of a comma-separated list of numbers and ranges (N-M means ports N through M, inclusive). It is formatted according to the following regular expression: "\d+(\-\d+)?(,\d+(\-\d+)?)*". For example, "2,5-15,30,32,40-50,55-60".
The PORTLIST data type is implemented as an "xs:string" with a regular expression constraint in the schema.
A postal address is represented by the POSTAL data type. This data type is an ML_STRING whose format is documented in Section 2.23 of RFC 4519 [10]. It defines a postal address as a free-form multi-line string separated by the "$" character.
The POSTAL data type is implemented as an "xs:string" in the schema.
The name of an individual or organization is represented by the NAME data type. This data type is an ML_STRING whose format is documented in Section 2.3 of RFC 4519 [10].
The NAME data type is implemented as an "xs:string" in the schema.
In this section, the individual components of the IODEF data model will be discussed in detail. For each class, the semantics will be described and the relationship with other classes will be depicted with UML. When necessary, specific comments will be made about corresponding definition in the schema in Section 8
The IODEF-Document class is the top level class in the IODEF data model. All IODEF documents are an instance of this class.
+-----------------+ | IODEF-Document | +-----------------+ | STRING version |<>--{1..*}--[ Incident ] | ENUM lang | | STRING formatid | +-----------------+
Figure 1: IODEF-Document Class
The aggregate class that constitute IODEF-Document is:
Incident One or more. The information related to a single incident.
The IODEF-Document class has three attributes:
version Required. STRING. The IODEF specification version number to which this IODEF document conforms. The value of this attribute MUST be "1.00"
lang Required. ENUM. A valid language code per RFC 4646 [7] constrained by the definition of "xs:language". The interpretation of this code is described in Section 6.
formatid Optional. STRING. A free-form string to convey processing instructions to the recipient of the document. Its semantics must be negotiated out-of-band.
Every incident is represented by an instance of the Incident class. This class provides a standardized representation for commonly exchanged incident data.
The aggregate classes that constitute Incident are:
IncidentID One. An incident tracking number assigned to this incident by the CSIRT that generated the IODEF document.
AlternativeID Zero or one. The incident tracking numbers used by other CSIRTs to refer to the incident described in the document.
RelatedActivity Zero or one. The incident tracking numbers of related incidents.
DetectTime Zero or one. The time the incident was first detected.
StartTime Zero or one. The time the incident started.
EndTime Zero or one. The time the incident ended.
ReportTime One. The time the incident was reported.
Danyliw, et al. Standards Track [Page 11]
RFC 5070 IODEF December 2007
Description Zero or more. ML_STRING. A free-form textual description of the incident.
Assessment One or more. A characterization of the impact of the incident.
Method Zero or more. The techniques used by the intruder in the incident.
Contact One or more. Contact information for the parties involved in the incident.
EventData Zero or more. Description of the events comprising the incident.
History Zero or one. A log of significant events or actions that occurred during the course of handling the incident.
AdditionalData Zero or more. Mechanism by which to extend the data model.
The Incident class has four attributes:
purpose Required. ENUM. The purpose attribute represents the reason why the IODEF document was created. It is closely related to the Expectation class (Section 3.13). This attribute is defined as an enumerated list:
1. traceback. The document was sent for trace-back purposes.
2. mitigation. The document was sent to request aid in mitigating the described activity.
3. reporting. The document was sent to comply with reporting requirements.
4. other. The document was sent for purposes specified in the Expectation class.
5. ext-value. An escape value used to extend this attribute. See Section 5.1.
Danyliw, et al. Standards Track [Page 12]
RFC 5070 IODEF December 2007
ext-purpose Optional. STRING. A means by which to extend the purpose attribute. See Section 5.1.
lang Optional. ENUM. A valid language code per RFC 4646 [7] constrained by the definition of "xs:language". The interpretation of this code is described in Section 6.
restriction Optional. ENUM. This attribute indicates the disclosure guidelines to which the sender expects the recipient to adhere for the information represented in this class and its children. This guideline provides no security since there are no specified technical means to ensure that the recipient of the document handles the information as the sender requested.
The value of this attribute is logically inherited by the children of this class. That is to say, the disclosure rules applied to this class, also apply to its children.
It is possible to set a granular disclosure policy, since all of the high-level classes (i.e., children of the Incident class) have a restriction attribute. Therefore, a child can override the guidelines of a parent class, be it to restrict or relax the disclosure rules (e.g., a child has a weaker policy than an ancestor; or an ancestor has a weak policy, and the children selectively apply more rigid controls). The implicit value of the restriction attribute for a class that did not specify one can be found in the closest ancestor that did specify a value.
This attribute is defined as an enumerated value with a default value of "private". Note that the default value of the restriction attribute is only defined in the context of the Incident class. In other classes where this attribute is used, no default is specified.
1. public. There are no restrictions placed in the information.
2. need-to-know. The information may be shared with other parties that are involved in the incident as determined by the recipient of this document (e.g., multiple victim sites can be informed of each other).
3. private. The information may not be shared.
Danyliw, et al. Standards Track [Page 13]
RFC 5070 IODEF December 2007
4. default. The information can be shared according to an information disclosure policy pre-arranged by the communicating parties.
The IncidentID class represents an incident tracking number that is unique in the context of the CSIRT and identifies the activity characterized in an IODEF Document. This identifier would serve as an index into the CSIRT incident handling system. The combination of the name attribute and the string in the element content MUST be a globally unique identifier describing the activity. Documents generated by a given CSIRT MUST NOT reuse the same value unless they are referencing the same incident.
name Required. STRING. An identifier describing the CSIRT that created the document. In order to have a globally unique CSIRT name, the fully qualified domain name associated with the CSIRT MUST be used.
instance Optional. STRING. An identifier referencing a subset of the named incident.
restriction Optional. ENUM. This attribute has been defined in Section 3.2.
The AlternativeID class lists the incident tracking numbers used by CSIRTs, other than the one generating the document, to refer to the identical activity described the IODEF document. A tracking number listed as an AlternativeID references the same incident detected by
Danyliw, et al. Standards Track [Page 14]
RFC 5070 IODEF December 2007
another CSIRT. The incident tracking numbers of the CSIRT that generated the IODEF document should never be considered an AlternativeID.
The RelatedActivity class lists either incident tracking numbers of incidents or URLs (not both) that refer to activity related to the one described in the IODEF document. These references may be to local incident tracking numbers or to those of other CSIRTs.
The specifics of how a CSIRT comes to believe that two incidents are related are considered out of scope.
The AdditionalData class serves as an extension mechanism for information not otherwise represented in the data model. For relatively simple information, atomic data types (e.g., integers, strings) are provided with a mechanism to annotate their meaning. The class can also be used to extend the data model (and the associated Schema) to support proprietary extensions by encapsulating entire XML documents conforming to another Schema (e.g., IDMEF). A detailed discussion for extending the data model and the schema can be found in Section 5.
Unlike XML, which is self-describing, atomic data must be documented to convey its meaning. This information is described in the 'meaning' attribute. Since these description are outside the scope of the specification, some additional coordination may be required to ensure that a recipient of a document using the AdditionalData classes can make sense of the custom extensions.
The Contact class describes contact information for organizations and personnel involved in the incident. This class allows for the naming of the involved party, specifying contact information for them, and identifying their role in the incident.
People and organizations are treated interchangeably as contacts; one can be associated with the other using the recursive definition of the class (the Contact class is aggregated into the Contact class). The 'type' attribute disambiguates the type of contact information being provided.
The inheriting definition of Contact provides a way to relate information without requiring the explicit use of identifiers in the classes or duplication of data. A complete point of contact is derived by a particular traversal from the root Contact class to the leaf Contact class. As such, multiple points of contact might be specified in a single instance of a Contact class. Each child Contact class logically inherits contact information from its ancestors.
The aggregate classes that constitute the Contact class are:
ContactName Zero or one. ML_STRING. The name of the contact. The contact may either be an organization or a person. The type attribute disambiguates the semantics.
Description Zero or many. ML_STRING. A free-form description of this contact. In the case of a person, this is often the organizational title of the individual.
RegistryHandle Zero or many. A handle name into the registry of the contact.
PostalAddress Zero or one. The postal address of the contact.
Email Zero or many. The email address of the contact.
Telephone Zero or many. The telephone number of the contact.
Fax Zero or one. The facsimile telephone number of the contact.
Timezone Zero or one. TIMEZONE. The timezone in which the contact resides formatted according to Section 2.9.
Danyliw, et al. Standards Track [Page 19]
RFC 5070 IODEF December 2007
Contact Zero or many. A Contact instance contained within another Contact instance inherits the values of the parent(s). This recursive definition can be used to group common data pertaining to multiple points of contact and is especially useful when listing multiple contacts at the same organization.
AdditionalData Zero or many. A mechanism by which to extend the data model.
At least one of the aggregate classes MUST be present in an instance of the Contact class. This is not enforced in the IODEF schema as there is no simple way to accomplish it.
The Contact class has five attributes:
role Required. ENUM. Indicates the role the contact fulfills. This attribute is defined as an enumerated list:
1. creator. The entity that generate the document.
2. admin. An administrative contact for a host or network.
3. tech. A technical contact for a host or network.
4. irt. The CSIRT involved in handling the incident.
5. cc. An entity that is to be kept informed about the handling of the incident.
6. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-role Optional. STRING. A means by which to extend the role attribute. See Section 5.1.
type Required. ENUM. Indicates the type of contact being described. This attribute is defined as an enumerated list:
1. person. The information for this contact references an individual.
2. organization. The information for this contact references an organization.
Danyliw, et al. Standards Track [Page 20]
RFC 5070 IODEF December 2007
3. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-type Optional. STRING. A means by which to extend the type attribute. See Section 5.1.
restriction Optional. ENUM. This attribute is defined in Section 3.2.
The RegistryHandle class represents a handle into an Internet registry or community-specific database. The handle is specified in the element content and the type attribute specifies the database.
The PostalAddress class specifies a postal address formatted according to the POSTAL data type (Section 2.11).
+---------------------+ | PostalAddress | +---------------------+ | POSTAL | | | | ENUM meaning | | ENUM lang | +---------------------+
Figure 9: The PostalAddress Class
The PostalAddress class has two attributes:
meaning Optional. ENUM. A free-form description of the element content.
lang Required. ENUM. A valid language code per RFC 4646 [7] constrained by the definition of "xs:language". The interpretation of this code is described in Section 6.
The data model uses five different classes to represent a timestamp. Their definition is identical, but each has a distinct name to convey a difference in semantics.
The element content of each class is a timestamp formatted according to the DATETIME data type (see Section 2.8).
The ReportTime class represents the time the incident was reported. This timestamp SHOULD coincide to the time at which the IODEF document is generated.
The Method class describes the methodology used by the intruder to perpetrate the events of the incident. This class consists of a list of references describing the attack method and a free form description of the technique.
The Reference class is a reference to a vulnerability, IDS alert, malware sample, advisory, or attack technique. A reference consists of a name, a URL to this reference, and an optional description.
The aggregate classes that constitute Assessment are:
Impact Zero or many. Technical impact of the incident on a network.
TimeImpact Zero or many. Impact of the activity measured with respect to time.
MonetaryImpact Zero or many. Impact of the activity measured with respect to financial loss.
Counter Zero or more. A counter with which to summarize the magnitude of the activity.
Confidence Zero or one. An estimate of confidence in the assessment.
AdditionalData Zero or many. A mechanism by which to extend the data model.
A least one instance of the possible three impact classes (i.e., Impact, TimeImpact, or MonetaryImpact) MUST be present.
The Assessment class has two attributes:
occurrence Optional. ENUM. Specifies whether the assessment is describing actual or potential outcomes. The default is "actual" and is assumed if not specified.
Danyliw, et al. Standards Track [Page 26]
RFC 5070 IODEF December 2007
1. actual. This assessment describes activity that has occurred.
2. potential. This assessment describes potential activity that might occur.
restriction Optional. ENUM. This attribute is defined in Section 3.2.
The element content will be a free-form textual description of the impact.
The Impact class has five attributes:
lang Required. ENUM. A valid language code per RFC 4646 [7] constrained by the definition of "xs:language". The interpretation of this code is described in Section 6.
severity Optional. ENUM. An estimate of the relative severity of the activity. The permitted values are shown below. There is no default value.
Danyliw, et al. Standards Track [Page 27]
RFC 5070 IODEF December 2007
1. low. Low severity
2. medium. Medium severity
3. high. High severity
completion Optional. ENUM. An indication whether the described activity was successful. The permitted values are shown below. There is no default value.
1. failed. The attempted activity was not successful.
2. succeeded. The attempted activity succeeded.
type Required. ENUM. Classifies the malicious activity into incident categories. The permitted values are shown below. The default value is "other".
1. admin. Administrative privileges were attempted.
2. dos. A denial of service was attempted.
3. file. An action that impacts the integrity of a file or database was attempted.
4. info-leak. An attempt was made to exfiltrate information.
5. misconfiguration. An attempt was made to exploit a mis- configuration in a system.
6. policy. Activity violating site's policy was attempted.
7. recon. Reconnaissance activity was attempted.
8. social-engineering. A social engineering attack was attempted.
9. user. User privileges were attempted.
10. unknown. The classification of this activity is unknown.
11. ext-value. An escape value used to extend this attribute. See Section 5.1.
Danyliw, et al. Standards Track [Page 28]
RFC 5070 IODEF December 2007
ext-type Optional. STRING. A means by which to extend the type attribute. See Section 5.1.
The TimeImpact class describes the impact of the incident on an organization as a function of time. It provides a way to convey down time and recovery time.
The element content is a positive, floating point (REAL) number specifying a unit of time. The duration and metric attributes will imply the semantics of the element content.
The TimeImpact class has five attributes:
severity Optional. ENUM. An estimate of the relative severity of the activity. The permitted values are shown below. There is no default value.
1. low. Low severity
2. medium. Medium severity
3. high. High severity
metric Required. ENUM. Defines the metric in which the time is expressed. The permitted values are shown below. There is no default value.
Danyliw, et al. Standards Track [Page 29]
RFC 5070 IODEF December 2007
1. labor. Total staff-time to recovery from the activity (e.g., 2 employees working 4 hours each would be 8 hours).
2. elapsed. Elapsed time from the beginning of the recovery to its completion (i.e., wall-clock time).
3. downtime. Duration of time for which some provided service(s) was not available.
4. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-metric Optional. STRING. A means by which to extend the metric attribute. See Section 5.1.
duration Required. ENUM. Defines a unit of time, that when combined with the metric attribute, fully describes a metric of impact that will be conveyed in the element content. The permitted values are shown below. The default value is "hour".
1. second. The unit of the element content is seconds.
2. minute. The unit of the element content is minutes.
3. hour. The unit of the element content is hours.
4. day. The unit of the element content is days.
5. month. The unit of the element content is months.
6. quarter. The unit of the element content is quarters.
7. year. The unit of the element content is years.
8. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-duration Optional. STRING. A means by which to extend the duration attribute. See Section 5.1.
The MonetaryImpact class describes the financial impact of the activity on an organization. For example, this impact may consider losses due to the cost of the investigation or recovery, diminished
Danyliw, et al. Standards Track [Page 30]
RFC 5070 IODEF December 2007
productivity of the staff, or a tarnished reputation that will affect future opportunities.
The element content is a positive, floating point number (REAL) specifying a unit of currency described in the currency attribute.
The MonetaryImpact class has two attributes:
severity Optional. ENUM. An estimate of the relative severity of the activity. The permitted values are shown below. There is no default value.
1. low. Low severity
2. medium. Medium severity
3. high. High severity
currency Required. STRING. Defines the currency in which the monetary impact is expressed. The permitted values are defined in ISO 4217:2001, Codes for the representation of currencies and funds [14]. There is no default value.
The Confidence class represents a best estimate of the validity and accuracy of the described impact (see Section 3.10) of the incident activity. This estimate can be expressed as a category or a numeric calculation.
The element content expresses a numerical assessment in the confidence of the data when the value of the rating attribute is "numeric". Otherwise, this element should be empty.
The Confidence class has one attribute.
rating Required. ENUM. A rating of the analytical validity of the specified Assessment. The permitted values are shown below. There is no default value.
1. low. Low confidence in the validity.
2. medium. Medium confidence in the validity.
3. high. High confidence in the validity.
4. numeric. The element content contains a number that conveys the confidence of the data. The semantics of this number outside the scope of this specification.
The HistoryItem class is an entry in the History (Section 3.11) log that documents a particular action or event that occurred in the course of handling the incident. The details of the entry are a free-form description, but each can be categorized with the type attribute.
The aggregate classes that constitute HistoryItem are:
DateTime One. Timestamp of this entry in the history log (e.g., when the action described in the Description was taken).
IncidentID Zero or One. In a history log created by multiple parties, the IncidentID provides a mechanism to specify which CSIRT created a particular entry and references this organization's incident tracking number. When a single organization is maintaining the log, this class can be ignored.
Danyliw, et al. Standards Track [Page 33]
RFC 5070 IODEF December 2007
Contact Zero or One. Provides contact information for the person that performed the action documented in this class.
Description Zero or many. ML_STRING. A free-form textual description of the action or event.
AdditionalData Zero or many. A mechanism by which to extend the data model.
The HistoryItem class has three attributes:
restriction Optional. ENUM. This attribute has been defined in Section 3.2.
action Required. ENUM. Classifies a performed action or occurrence documented in this history log entry. As activity will likely have been instigated either through a previously conveyed expectation or internal investigation, this attribute is identical to the category attribute of the Expectation class. The difference is only one of tense. When an action is in this class, it has been completed. See Section 3.13.
ext-action Optional. STRING. A means by which to extend the action attribute. See Section 5.1.
The EventData class describes a particular event of the incident for a given set of hosts or networks. This description includes the systems from which the activity originated and those targeted, an assessment of the techniques used by the intruder, the impact of the activity on the organization, and any forensic evidence discovered.
The aggregate classes that constitute EventData are:
Description Zero or more. ML_STRING. A free-form textual description of the event.
DetectTime Zero or one. The time the event was detected.
StartTime Zero or one. The time the event started.
EndTime Zero or one. The time the event ended.
Contact Zero or more. Contact information for the parties involved in the event.
Assessment Zero or one. The impact of the event on the target and the actions taken.
Method Zero or more. The technique used by the intruder in the event.
Danyliw, et al. Standards Track [Page 35]
RFC 5070 IODEF December 2007
Flow Zero or more. A description of the systems or networks involved.
Expectation Zero or more. The expected action to be performed by the recipient for the described event.
Record Zero or one. Supportive data (e.g., log files) that provides additional information about the event.
EventData Zero or more. EventData instances contained within another EventData instance inherit the values of the parent(s); this recursive definition can be used to group common data pertaining to multiple events. When EventData elements are defined recursively, only the leaf instances (those EventData instances not containing other EventData instances) represent actual events.
AdditionalData Zero or more. An extension mechanism for data not explicitly represented in the data model.
At least one of the aggregate classes MUST be present in an instance of the EventData class. This is not enforced in the IODEF schema as there is no simple way to accomplish it.
The EventData class has one attribute:
restriction Optional. ENUM. This attribute is defined in Section 3.2.
3.12.1. Relating the Incident and EventData Classes
There is substantial overlap in the Incident and EventData classes. Nevertheless, the semantics of these classes are quite different. The Incident class provides summary information about the entire incident, while the EventData class provides information about the individual events comprising the incident. In the most common case, the EventData class will provide more specific information for the general description provided in the Incident class. However, it may also be possible that the overall summarized information about the incident conflicts with some individual information in an EventData class when there is a substantial composition of various events in the incident. In such a case, the interpretation of the more specific EventData MUST supersede the more generic information provided in IncidentData.
The EventData class can be thought of as a container for the properties of an event in an incident. These properties include: the hosts involved, impact of the incident activity on the hosts, forensic logs, etc. With an instance of the EventData class, hosts (i.e., System class) are grouped around these common properties.
The recursive definition (or instance property inheritance) of the EventData class (the EventData class is aggregated into the EventData class) provides a way to related information without requiring the explicit use of unique attribute identifiers in the classes or duplicating information. Instead, the relative depth (nesting) of a class is used to group (relate) information.
For example, an EventData class might be used to describe two machines involved in an incident. This description can be achieved using multiple instances of the Flow class. It happens that there is a common technical contact (i.e., Contact class) for these two machines, but the impact (i.e., Assessment class) on them is different. A depiction of the representation for this situation can be found in Figure 23.
The Expectation class conveys to the recipient of the IODEF document the actions the sender is requesting. The scope of the requested action is limited to purview of the EventData class in which this class is aggregated.
The aggregate classes that constitute Expectation are:
Description Zero or many. ML_STRING. A free-form description of the desired action(s).
StartTime Zero or one. The time at which the action should be performed. A timestamp that is earlier than the ReportTime specified in the Incident class denotes that the expectation should be fulfilled as soon as possible. The absence of this element leaves the execution of the expectation to the discretion of the recipient.
EndTime Zero or one. The time by which the action should be completed. If the action is not carried out by this time, it should no longer be performed.
Contact Zero or one. The expected actor for the action.
The Expectations class has four attributes:
restriction Optional. ENUM. This attribute is defined in Section 3.2.
severity Optional. ENUM. Indicates the desired priority of the action. This attribute is an enumerated list with no default value, and the semantics of these relative measures are context dependent.
1. low. Low priority
2. medium. Medium priority
3. high. High priority
Danyliw, et al. Standards Track [Page 38]
RFC 5070 IODEF December 2007
action Optional. ENUM. Classifies the type of action requested. This attribute is an enumerated list with no default value.
1. nothing. No action is requested. Do nothing with the information.
2. contact-source-site. Contact the site(s) identified as the source of the activity.
3. contact-target-site. Contact the site(s) identified as the target of the activity.
4. contact-sender. Contact the originator of the document.
5. investigate. Investigate the systems(s) listed in the event.
6. block-host. Block traffic from the machine(s) listed as sources the event.
7. block-network. Block traffic from the network(s) lists as sources in the event.
8. block-port. Block the port listed as sources in the event.
9. rate-limit-host. Rate-limit the traffic from the machine(s) listed as sources in the event.
10. rate-limit-network. Rate-limit the traffic from the network(s) lists as sources in the event.
11. rate-limit-port. Rate-limit the port(s) listed as sources in the event.
12. remediate-other. Remediate the activity in a way other than by rate limiting or blocking.
13. status-triage. Conveys receipts and the triaging of an incident.
14. status-new-info. Conveys that new information was received for this incident.
15. other. Perform some custom action described in the Description class.
16. ext-value. An escape value used to extend this attribute. See Section 5.1.
Danyliw, et al. Standards Track [Page 39]
RFC 5070 IODEF December 2007
ext-action Optional. STRING. A means by which to extend the action attribute. See Section 5.1.
The System class describes a system or network involved in an event. The systems or networks represented by this class are categorized according to the role they played in the incident through the category attribute. The value of this category attribute dictates the semantics of the aggregated classes in the System class. If the category attribute has a value of "source", then the aggregated classes denote the machine and service from which the activity is originating. With a category attribute value of "target" or "intermediary", then the machine or service is the one targeted in the activity. A value of "sensor" dictates that this System was part of an instrumentation to monitor the network.
Node One. A host or network involved in the incident.
Service Zero or more. A network service running on the system.
OperatingSystem Zero or one. The operating system running on the system.
Counter Zero or more. A counter with which to summarize properties of this host or network.
Description Zero or more. ML_STRING. A free-form text description of the System.
AdditionalData Zero or many. A mechanism by which to extend the data model.
The System class has five attributes:
restriction Optional. ENUM. This attribute is defined in Section 3.2.
category Required. ENUM. Classifies the role the host or network played in the incident. The possible values are:
1. source. The System was the source of the event.
2. target. The System was the target of the event.
Danyliw, et al. Standards Track [Page 41]
RFC 5070 IODEF December 2007
3. intermediate. The System was an intermediary in the event.
4. sensor. The System was a sensor monitoring the event.
5. infrastructure. The System was an infrastructure node of IODEF document exchange.
6. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-category Optional. STRING. A means by which to extend the category attribute. See Section 5.1.
interface Optional. STRING. Specifies the interface on which the event(s) on this System originated. If the Node class specifies a network rather than a host, this attribute has no meaning.
spoofed Optional. ENUM. An indication of confidence in whether this System was the true target or attacking host. The permitted values for this attribute are shown below. The default value is "unknown".
1. unknown. The accuracy of the category attribute value is unknown.
2. yes. The category attribute value is probably incorrect. In the case of a source, the System is likely a decoy; with a target, the System was likely not the intended victim.
3. no. The category attribute value is believed to be correct.
NodeName Zero or more. ML_STRING. The name of the Node (e.g., fully qualified domain name). This information MUST be provided if no Address information is given.
Address Zero or more. The hardware, network, or application address of the Node. If a NodeName is not provided, at least one Address MUST be specified.
Location Zero or one. ML_STRING. A free-from description of the physical location of the equipment.
DateTime Zero or one. A timestamp of when the resolution between the name and address was performed. This information SHOULD be provided if both an Address and NodeName are specified.
NodeRole Zero or more. The intended purpose of the Node.
Counter Zero or more. A counter with which to summarizes properties of this host or network.
The Counter class summarize multiple occurrences of some event, or conveys counts or rates on various features (e.g., packets, sessions, events).
Danyliw, et al. Standards Track [Page 43]
RFC 5070 IODEF December 2007
The value of the counter is the element content with its units represented in the type attribute. A rate for a given feature can be expressed by setting the duration attribute. The complete semantics are entirely context dependent based on the class in which the Counter is aggregated.
type Required. ENUM. Specifies the units of the element content.
1. byte. Count of bytes.
2. packet. Count of packets.
3. flow. Count of flow (e.g., NetFlow records).
4. session. Count of sessions.
5. alert. Count of notifications generated by another system (e.g., IDS or SIM).
6. message. Count of messages (e.g., mail messages).
7. event. Count of events.
8. host. Count of hosts.
9. site. Count of site.
10. organization. Count of organizations.
Danyliw, et al. Standards Track [Page 44]
RFC 5070 IODEF December 2007
11. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-type Optional. STRING. A means by which to extend the type attribute. See Section 5.1.
duration Optional. ENUM. If present, the Counter class represents a rate rather than a count over the entire event. In that case, this attribute specifies the denominator of the rate (where the type attribute specified the nominator). The possible values of this attribute are defined in Section 3.10.2
ext-duration Optional. STRING. A means by which to extend the duration attribute. See Section 5.1.
category Required. ENUM. Functionality provided by a node.
1. client. Client computer
2. server-internal. Server with internal services
3. server-public. Server with public services
4. www. WWW server
5. mail. Mail server
6. messaging. Messaging server (e.g., NNTP, IRC, IM)
7. streaming. Streaming-media server
8. voice. Voice server (e.g., SIP, H.323)
9. file. File server (e.g., SMB, CVS, AFS)
10. ftp. FTP server
11. p2p. Peer-to-peer node
12. name. Name server (e.g., DNS, WINS)
13. directory. Directory server (e.g., LDAP, finger, whois)
14. credential. Credential server (e.g., domain controller, Kerberos)
15. print. Print server
16. application. Application server
Danyliw, et al. Standards Track [Page 47]
RFC 5070 IODEF December 2007
17. database. Database server
18. infra. Infrastructure server (e.g., router, firewall, DHCP)
19. log. Logserver (e.g., syslog)
20. ext-value. An escape value used to extend this attribute. See Section 5.1.
ext-category Optional. STRING. A means by which to extend the category attribute. See Section 5.1.
lang Required. ENUM. A valid language code per RFC 4646 [7] constrained by the definition of "xs:language". The interpretation of this code is described in Section 6.
The Service class describes a network service of a host or network. The service is identified by specific port or list of ports, along with the application listening on that port.
When Service occurs as an aggregate class of a System that is a source, then this service is the one from which activity of interest is originating. Conversely, when Service occurs as an aggregate class of a System that is a target, then that service is the one to which activity of interest is directed.
The aggregate classes that constitute Service are:
Danyliw, et al. Standards Track [Page 48]
RFC 5070 IODEF December 2007
Port Zero or one. INTEGER. A port number.
Portlist Zero or one. PORTLIST. A list of port numbers formatted according to Section 2.10.
ProtoCode Zero or one. INTEGER. A layer-4 protocol-specific code field (e.g., ICMP code field).
ProtoType Zero or one. INTEGER. A layer-4 protocol specific type field (e.g., ICMP type field).
ProtoFlags Zero or one. INTEGER. A layer-4 protocol specific flag field (e.g., TCP flag field).
Application Zero or more. The application bound to the specified Port or Portlist.
Either a Port or Portlist class MUST be specified for a given instance of a Service class.
For a given source, System@type="source", a corresponding target, System@type="target", maybe defined, or vice versa. When a Portlist class is defined in the Service class of both the source and target in a given instance of the Flow class, there MUST be symmetry in the enumeration of the ports. Thus, if n-ports are listed for a source, n-ports should be listed for the target. Likewise, the ports should be listed in an identical sequence such that the n-th port in the source corresponds to the n-th port of the target. This symmetry in listing and sequencing of ports applies whether there are 1-to-1, 1-to-many, or many-to-many sources-to-targets. In the 1-to-many or many-to-many, the exact order in which the System classes are enumerated in the Flow class is significant.
The Service class has one attribute:
ip_protocol Required. INTEGER. The IANA protocol number.
The Record class is a container class for log and audit data that provides supportive information about the incident. The source of this data will often be the output of monitoring tools. These logs should substantiate the activity described in the document.
RecordData One or more. Log or audit data generated by a particular type of sensor. Separate instances of the RecordData class SHOULD be used for each sensor type.
The Record class has one attribute:
restriction Optional. ENUM. This attribute has been defined in Section 3.2.
The aggregate classes that constitutes RecordData is:
DateTime Zero or one. Timestamp of the RecordItem data.
Description Zero or more. ML_STRING. Free-form textual description of the provided RecordItem data. At minimum, this description should convey the significance of the provided RecordItem data.
Application Zero or one. Information about the sensor used to generate the RecordItem data.
RecordPattern Zero or more. A search string to precisely find the relevant data in a RecordItem.
RecordItem One or more. Log, audit, or forensic data.
AdditionalData Zero or one. An extension mechanism for data not explicitly represented in the data model.
The RecordData class has one attribute:
restriction Optional. ENUM. This attribute has been defined in Section 3.2.
The RecordPattern class describes where in the content of the RecordItem relevant information can be found. It provides a way to reference subsets of information, identified by a pattern, in a large log file, audit trail, or forensic data.
The RecordItem class provides a way to incorporate relevant logs, audit trails, or forensic data to support the conclusions made during the course of analyzing the incident. The class supports both the direct encapsulation of the data, as well as, provides primitives to reference data stored elsewhere.
This class is identical to AdditionalData class (Section 3.6).
4. Processing Considerations
This section defines additional requirements on creating and parsing IODEF documents.
Every IODEF document MUST begin with an XML declaration, and MUST specify the XML version used. If UTF-8 encoding is not used, the character encoding MUST also be explicitly specified. The IODEF conforms to all XML data encoding conventions and constraints.
The XML declaration with no character encoding will read as follows:
<?xml version="1.0" ?>
When a character encoding is specified, the XML declaration will read like the following:
Danyliw, et al. Standards Track [Page 54]
RFC 5070 IODEF December 2007
<?xml version="1.0" encoding="charset" ?>
Where "charset" is the name of the character encoding as registered with the Internet Assigned Numbers Authority (IANA), see [9].
The following characters have special meaning in XML and MUST be escaped with their entity reference equivalent: "&", "<", ">", "\"" (double quotation mark), and "'" (apostrophe). These entity references are "&", "<", ">", """, and "'" respectively.
The IODEF schema declares a namespace of "urn:ietf:params:xml:ns:iodef-1.0" and registers it per [4]. Each IODEF document SHOULD include a valid reference to the IODEF schema using the "xsi:schemaLocation" attribute. An example of such a declaration would look as follows:
The IODEF documents MUST be well-formed XML and SHOULD be validated against the schema described in Section 8. However, mere conformance to the schema is not sufficient for a semantically valid IODEF document. There is additional specification in the text of Section 3 that cannot be readily encoded in the schema and it must also be considered by an IODEF parser. The following is a list of discrepancies in what is more strictly specified in the normative text (Section 3), but not enforced in the IODEF schema:
o The elements or attributes that are defined as POSTAL, NAME, PHONE, and EMAIL data-types are implemented as "xs:string", but more rigid formatting requirements are specified in the text.
o The IODEF-Document@lang and MLStringType@lang attributes are declared as an "xs:language" that constrains values with a regular expression. However, the value of this attribute still needs to be validated against the list of possible enumerated values is defined in [7].
o The MonetaryImpact@currency attribute is declared as an "xs: string", but the list of valid values as defined in [14].
Danyliw, et al. Standards Track [Page 55]
RFC 5070 IODEF December 2007
o All of the aggregated classes Contact and EventData are optional in the schema, but at least one of these aggregated classes MUST be present.
o There are multiple conventions that can be used to categorize a system using the NodeRole class or to specify software with the Application and OperatingSystem classes. IODEF parsers MUST accept incident reports that do not use these fields in accordance with local conventions.
o The Confidence@rating attribute determines whether the element content of Confidence should be empty.
o The Address@type attribute determines the format of the element content.
o The attributes AdditionalData@dtype and RecordItem@dtype derived from iodef:ExtensionType determine the semantics and formatting of the element content.
o Symmetry in the enumerated ports of a Portlist class is required between sources and targets. See Section 3.17.
5. Extending the IODEF
In order to support the changing activity of CSIRTS, the IODEF data model will need to evolve along with them. This section discusses how new data elements that have no current representation in the data model can be incorporated into the IODEF. These techniques are designed so that adding new data will not require a change to the IODEF schema. With proven value, well documented extensions can be incorporated into future versions of the specification. However, this approach also supports private extensions relevant only to a closed consortium.
5.1. Extending the Enumerated Values of Attributes
The data model supports a means by which to add new enumerated values to an attribute. For each attribute that supports this extension technique, there is a corresponding attribute in the same element whose name is identical, less a prefix of "ext-". This special attribute is referred to as the extension attribute, and the attribute being extended is referred to as an extensible attribute. For example, an extensible attribute named "foo" will have a corresponding extension attribute named "ext-foo". An element may have many extensible, and therefore many extension, attributes.
Danyliw, et al. Standards Track [Page 56]
RFC 5070 IODEF December 2007
In addition to a corresponding extension attribute, each extensible attribute has "ext-value" as one its possible values. This particular value serves as an escape sequence and has no valid meaning.
In order to add a new enumerated value to an extensible attribute, the value of this attribute MUST be set to "ext-value", and the new desired value MUST be set in the corresponding extension attribute. For example, an extended instance of the type attribute of the Impact class would look as follows:
The classes of the data model can be extended only through the use of the AdditionalData and RecordItem classes. These container classes, collectively referred to as the extensible classes, are implemented with the iodef:ExtensionType data type in the schema. They provide the ability to have new atomic or XML-encoded data elements in all of the top-level classes of the Incident class and a few of the more complicated subordinate classes. As there are multiple instances of the extensible classes in the data model, there is discretion on where to add a new data element. It is RECOMMENDED that the extension be placed in the most closely related class to the new information.
Extensions using the atomic data types (i.e., all values of the dtype attributes other than "xml") MUST:
1. Set the element content of extensible class to the desired value, and
2. Set the dtype attribute to correspond to the data type of the element content.
The following guidelines exist for extensions using XML:
1. The element content of the extensible class MUST be set to the desired value and the dtype attribute MUST be set to "xml".
2. The extension schema MUST declare a separate namespace. It is RECOMMENDED that these extensions have the prefix "iodef-".
Danyliw, et al. Standards Track [Page 57]
RFC 5070 IODEF December 2007
3. It is RECOMMENDED that extension schemas follow the naming convention of the IODEF data model. The names of all elements are capitalized. For composed names, a capital letter is used for each word. Attribute names are lower case.
4. When a parser encounters an IODEF document with an extension it does not understand, this extension MUST be ignored (and not processed), but the remainder of the document MUST be processed. Parsers will be able to identify these extensions for which they have no processing logic through the namespace declaration. Parsers that encounter an unrecognized element in a namespace that they do support SHOULD reject the document as a syntax error.
5. Implementations SHOULD NOT download schemas at runtime due to the security implications, and extensions MUST NOT be required to provide a resolvable location of their schema.
The following schema and XML document excerpt provide a template for an extension schema and its use in the IODEF document.
This example schema defines a namespace of "iodef-extension1" and a single element named "newdata".
The following XML excerpt demonstrates the use of the above schema as an extension to the IODEF.
Danyliw, et al. Standards Track [Page 58]
RFC 5070 IODEF December 2007
<IODEF-Document version="1.00" lang="en-US" xmlns="urn:ietf:params:xml:ns:iodef-1.0" xmlns:iodef=" urn:ietf:params:xml:ns:iodef-1.0" xmlns:iodef-extension1="iodef-extension1.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="iodef-extension1.xsd"> <Incident purpose="reporting"> ... <AdditionalData dtype="xml" meaning="xml"> <iodef-extension1:newdata> Field that could not be represented elsewhere </iodef-extension1:newdata> </AdditionalData> </Incident> </IODEF-Document>
6. Internationalization Issues
Internationalization and localization is of specific concern to the IODEF, since it is only through collaboration, often across language barriers, that certain incidents be resolved. The IODEF supports this goal by depending on XML constructs, and through explicit design choices in the data model.
Since IODEF is implemented as an XML Schema, it implicitly supports all the different character encodings, such as UTF-8 and UTF-16, possible with XML. Additionally, each IODEF document MUST specify the language in which their contents are encoded. The language can be specified with the attribute "xml:lang" (per Section 2.12 of [1]) in the top-level element (i.e., IODEF-Document@lang) and letting all other elements inherit that definition. All IODEF classes with a free-form text definition (i.e., all those defined of type iodef: MLStringType) can also specify a language different from the rest of the document. The valid language codes for the "xml:lang" attribute are described in RFC 4646 [7].
The data model supports multiple translations of free-form text. In the places where free-text is used for descriptive purposes, the given class always has a one-to-many cardinality to its parent (e.g., Description class). The intent is to allow the identical text to be encoded in different instances of the same class, but each being in a different language. This approach allows an IODEF document author to send recipients speaking different languages an identical document. The IODEF parser SHOULD extract the appropriate language relevant to the recipient.
Danyliw, et al. Standards Track [Page 59]
RFC 5070 IODEF December 2007
While the intent of the data model is to provide internationalization and localization, the intent is not to do so at the detriment of interoperability. While the IODEF does support different languages, the data model also relies heavily on standardized enumerated attributes that can crudely approximate the contents of the document. With this approach, a CSIRT should be able to make some sense of an IODEF document it receives even if the text based data elements are written in a language unfamiliar to the analyst.
7. Examples
This section provides examples of an incident encoded in the IODEF. These examples do not necessarily represent the only way to encode a particular incident.
<?xml version="1.0" encoding="UTF-8" ?> <!-- This example describes a compromise and subsequent installation of bots --> <IODEF-Document version="1.00" lang="en" xmlns="urn:ietf:params:xml:ns:iodef-1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="urn:ietf:params:xml:schema:iodef-1.0"> <Incident purpose="mitigation"> <IncidentID name="csirt.example.com">908711</IncidentID> <ReportTime>2006-06-08T05:44:53-05:00</ReportTime> <Description>Large bot-net</Description> <Assessment> <Impact type="dos" severity="high" completion="succeeded" />
Danyliw, et al. Standards Track [Page 63]
RFC 5070 IODEF December 2007
</Assessment> <Method> <!-- References a given piece of malware, "GT Bot" --> <Reference> <ReferenceName>GT Bot</ReferenceName> </Reference> <!-- References the vulnerability used to compromise the machines --> <Reference> <ReferenceName>CA-2003-22</ReferenceName> <URL>http://www.cert.org/advisories/CA-2003-22.html</URL> <Description>Root compromise via this IE vulnerability to install the GT Bot</Description> </Reference> </Method> <!-- A member of the CSIRT that is coordinating this incident --> <Contact type="person" role="irt"> <ContactName>Joe Smith</ContactName> <Email>jsmith@csirt.example.com</Email> </Contact> <EventData> <Description>These hosts are compromised and acting as bots communicating with irc.example.com.</Description> <Flow> <!-- bot running on 192.0.2.1 and sending DoS traffic at 10,000 bytes/second --> <System category="source"> <Node> <Address category="ipv4-addr">192.0.2.1</Address> </Node> <Counter type="byte" duration="second">10000</Counter> <Description>bot</Description> </System> <!-- a second bot on 192.0.2.3 --> <System category="source"> <Node> <Address category="ipv4-addr">192.0.2.3</Address> </Node> <Counter type="byte" duration="second">250000</Counter> <Description>bot</Description> </System> <!-- Command-and-control IRC server for these bots--> <System category="intermediate"> <Node> <NodeName>irc.example.com</NodeName> <Address category="ipv4-addr">192.0.2.20</Address> <DateTime>2006-06-08T01:01:03-05:00</DateTime>
Danyliw, et al. Standards Track [Page 64]
RFC 5070 IODEF December 2007
</Node> <Description>IRC server on #give-me-cmd channel</Description> </System> </Flow> <!-- Request to take these machines offline --> <Expectation action="investigate"> <Description>Confirm the source and take machines off-line and remediate</Description> </Expectation> </EventData> </Incident> </IODEF-Document>
<?xml version="1.0" encoding="UTF-8" ?> <!-- This example demonstrates a trivial IP watch-list --> <!-- @formatid is set to "watch-list-043" to demonstrate how additional semantics about this document could be conveyed assuming both parties understood it--> <IODEF-Document version="1.00" lang="en" formatid="watch-list-043" xmlns="urn:ietf:params:xml:ns:iodef-1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="urn:ietf:params:xml:schema:iodef-1.0"> <Incident purpose="reporting" restriction="private"> <IncidentID name="csirt.example.com">908711</IncidentID> <ReportTime>2006-08-01T00:00:00-05:00</ReportTime> <Description>Watch-list of known bad IPs or networks</Description> <Assessment> <Impact type="admin" completion="succeeded" /> <Impact type="recon" completion="succeeded" /> </Assessment> <Contact type="organization" role="creator"> <ContactName>CSIRT for example.com</ContactName> <Email>contact@csirt.example.com</Email> </Contact> <!-- Separate <EventData> used to convey different <Expectation> --> <EventData> <Flow> <System category="source"> <Node> <Address category="ipv4-addr">192.0.2.53</Address> </Node>
Danyliw, et al. Standards Track [Page 65]
RFC 5070 IODEF December 2007
<Description>Source of numerous attacks</Description> </System> </Flow> <!-- Expectation class indicating that sender of list would like to be notified if activity from the host is seen --> <Expectation action="contact-sender" /> </EventData> <EventData> <Flow> <System category="source"> <Node> <Address category="ipv4-net">192.0.2.16/28</Address> </Node> <Description> Source of heavy scanning over past 1-month </Description> </System> </Flow> <Flow> <System category="source"> <Node> <Address category="ipv4-addr">192.0.2.241</Address> </Node> <Description>C2 IRC server</Description> </System> </Flow> <!-- Expectation class recommends that these networks be filtered --> <Expectation action="block-host" /> </EventData> </Incident> </IODEF-Document>
The IODEF data model itself does not directly introduce security issues. Rather, it simply defines a representation for incident information. As the data encoded by the IODEF might be considered privacy sensitive by the parties exchanging the information or by those described by it, care needs to be taken in ensuring the appropriate disclosure during both document exchange and subsequent processing. The former must be handled by a messaging format, but the latter risk must be addressed by the systems that process, store, and archive IODEF documents and information derived from them.
The contents of an IODEF document may include a request for action or an IODEF parser may independently have logic to take certain actions based on information that it finds. For this reason, care must be taken by the parser to properly authenticate the recipient of the document and ascribe an appropriate confidence to the data prior to action.
The underlying messaging format and protocol used to exchange instances of the IODEF MUST provide appropriate guarantees of confidentiality, integrity, and authenticity. The use of a standardized security protocol is encouraged. The Real-time Inter- network Defense (RID) protocol [18] and its associated transport binding IODEF/RID over SOAP [19] provide such security.
In order to suggest data processing and handling guidelines of the encoded information, the IODEF allows a document sender to convey a privacy policy using the restriction attribute. The various instances of this attribute allow different data elements of the document to be covered by dissimilar policies. While flexible, it must be stressed that this approach only serves as a guideline from the sender, as the recipient is free to ignore it. The issue of enforcement is not a technical problem.
Danyliw, et al. Standards Track [Page 87]
RFC 5070 IODEF December 2007
10. IANA Considerations
This document uses URNs to describe an XML namespace and schema conforming to a registry mechanism described in [15]
Registration for the IODEF namespace:
o URI: urn:ietf:params:xml:ns:iodef-1.0
o Registrant Contact: See the first author of the "Author's Address" section of this document.
o XML: None. Namespace URIs do not represent an XML specification.
Registration for the IODEF XML schema:
o URI: urn:ietf:params:xml:schema:iodef-1.0
o Registrant Contact: See the first author of the "Author's Address" section of this document.
o XML: See the "IODEF Schema" in Section 8 of this document.
11. Acknowledgments
The following groups and individuals, listed alphabetically, contributed substantially to this document and should be recognized for their efforts.
o Patrick Cain, Cooper-Cain Group, Inc.
o The eCSIRT.net Project
o The Incident Object Description and Exchange Format Working-Group of the TERENA task-force (TF-CSIRT)
o Glenn Mansfield Keeni, Cyber Solutions, Inc.
o Hiroyuki Kido, NARA Institute of Science and Technology
[1] World Wide Web Consortium, "Extensible Markup Language (XML) 1.0 (Second Edition)", W3C Recommendation , October 2000, <http://www.w3.org/TR/2000/REC-xml-20001006>.
[2] World Wide Web Consortium, "XML XML Schema Part 1: Structures Second Edition", W3C Recommendation , October 2004, <http://www.w3.org/TR/xmlschema-1/>.
[3] World Wide Web Consortium, "XML Schema Part 2: Datatypes Second Edition", W3C Recommendation , October 2004, <http://www.w3.org/TR/xmlschema-2/>.
[5] World Wide Web Consortium, "XML Path Language (XPath) 2.0", W3C Candidate Recommendation , June 2006, <http://www.w3.org/TR/xpath20/>.
[6] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", RFC 2119, March 1997.
[7] Philips, A. and M. Davis, "Tags for Identifying of Languages", RFC 4646, September 2006.
[8] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform Resource Identifiers (URI): Generic Syntax", RFC 3986, January 2005`.
[9] Freed, N. and J. Postel, "IANA Charset Registration Procedures", BCP 2978, October 2000.
[10] Sciberras, A., "Schema for User Applications", RFC 4519, June 2006.
[11] Resnick, P., "Internet Message Format", RFC 2822, April 2001.
[12] Klyne, G. and C. Newman, "Date and Time on the Internet: Timestamps", RFC 3339, July 2002.
Danyliw, et al. Standards Track [Page 89]
RFC 5070 IODEF December 2007
[13] International Organization for Standardization, "International Standard: Data elements and interchange formats - Information interchange - Representation of dates and times", ISO 8601, Second Edition, December 2000.
[14] International Organization for Standardization, "International Standard: Codes for the representation of currencies and funds, ISO 4217:2001", ISO 4217:2001, August 2001.
[15] Mealling, M., "The IETF XML Registry", RFC 3688, January 2004.
[16] Keeni, G., Demchenko, Y., and R. Danyliw, "Requirements for the Format for Incident Information Exchange (FINE)", Work in Progress, June 2006.
[17] Debar, H., Curry, D., Debar, H., and B. Feinstein, "Intrusion Detection Message Exchange Format", RFC 4765, March 2007.
[18] Moriarty, K., "Real-time Inter-network Defense", Work in Progress, April 2007.
[19] Moriarty, K. and B. Trammell, "IODEF/RID over SOAP", Work in Progress, April 2007.
[20] Shafranovich, Y., "Common Format and MIME Type for Comma- Separated Values (CSV) File", RFC 4180, October 2005.
Danyliw, et al. Standards Track [Page 90]
RFC 5070 IODEF December 2007
Authors' Addresses
Roman Danyliw CERT - Software Engineering Institute Pittsburgh, PA USA
EMail: rdd@cert.org
Jan Meijer
EMail: jan@flyingcloggies.nl
Yuri Demchenko University of Amsterdam Amsterdam Netherlands
EMail: demch@chello.nl
Danyliw, et al. Standards Track [Page 91]
RFC 5070 IODEF December 2007
Full Copyright Statement
Copyright (C) The IETF Trust (2007).
This document is subject to the rights, licenses and restrictions contained in BCP 78, and except as set forth therein, the authors retain all their rights.
This document and the information contained herein are provided on an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Intellectual Property
The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; nor does it represent that it has made any independent effort to identify any such rights. Information on the procedures with respect to rights in RFC documents can be found in BCP 78 and BCP 79.
Copies of IPR disclosures made to the IETF Secretariat and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this specification can be obtained from the IETF on-line IPR repository at http://www.ietf.org/ipr.
The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights that may cover technology that may be required to implement this standard. Please address the information to the IETF at ietf-ipr@ietf.org.