I2NSF Consumer-Facing Interface YANG Data Model
Department of Software
Sungkyunkwan University2066 Seobu-Ro, Jangan-GuSuwonGyeonggi-Do16419Republic of Korea+82 31 299 4957+82 31 290 7996pauljeong@skku.eduhttp://iotlab.skku.edu/people-jaehoon-jeong.php
Department of Electrical and Computer Engineering
Sungkyunkwan University2066 Seobu-Ro, Jangan-GuSuwonGyeonggi-Do16419Republic of Korea+82 31 299 4104eskim86@skku.eduhttp://seclab.skku.edu/people/eunsoo-kim/
Korea Telecom
70 Yuseong-Ro, Yuseong-GuDaejeon305-811Republic of Korea+82 42 870 8409taejin.ahn@kt.com
Juniper Networks
1133 Innovation WaySunnyvaleCA94089USArkkumar@juniper.net
Huawei
7453 Hickory HillSalineMI48176USA+1-734-604-0332shares@ndzh.com
Security
I2NSF Working GroupInternet-Draft
This document describes an information model and a YANG data model for
the Consumer-Facing Interface between an Interface to Network Security
Functions (I2NSF) User and Security Controller in an I2NSF system in
a Network Functions Virtualization (NFV) environment. The information
model defines various managed objects and relationship among these
objects needed to build the interface. The information model is
organized based on the "Event-condition-Event" (ECA) policy model
defined by a capability information model for Interface to Network
Security Functions (I2NSF),
and the data model is defined for enabling different users of a given
I2NSF system to define, manage, and monitor security policies for
specific flows within an administrative domain.
In an I2NSF framework, each vendor can register their NSFs using
a Developer's Management System (DMS). Assuming that vendors also provide
the front-end web applications registered with an I2NSF User,
the Consumer-Facing Interface is required because the web applications
developed by each vendor need to have a standard interface specifying
the data types used when the I2NSF User and Security Controller
communicate using this interface. Therefore, this document specifies
the required information, their data types, and encoding schemes
so that high-level security policies (or configuration information for
security policies) can be transferred to the Security Controller through
the Consumer-Facing Interface. These policies can easily be translated by the
Security Controller into low-level security policies.
The Security Controller delivers the translated policies to
Network Security Functions (NSFs) according to their respective
security capabilities for the required securiy enforcement.
The Consumer-Facing Interface would be built using a set of objects, with each
object capturing a unique set of information from Security Administrator
(i.e., I2NSF User ) needed to express a Security Policy.
An object may have relationship with various other objects to express a
complete set of requirement. An information model captures the managed objects
and relationship among these objects. The information model proposed in this
document is structured in accordance with the "Event-Condition-Event" (ECA)
policy model.
An NSF Capability model is proposed in as the basic model for both the NSF-Facing interface and Consumer-Facing Interface security policy model of this document.
explains differences between an information and data model.
This document use the guidelines in to define both the
information and data model for Consumer-Facing Interface.
shows a high-level abstraction
of Consumer-Facing Interface. A data model, which represents an implementation
of the information model in a specific data representation language, is also
defined in this document.
Data models are defined at a lower level of abstraction and provide many details.
They provide details about the implementation of a protocol’s specification,
e.g., rules that explain how to map managed objects onto lower-level protocol
constructs. Since conceptual models can be implemented in different ways,
multiple data models can be derived by a single information model.
The efficient and flexible provisioning of network functions
by a Network Functions Virtualization (NFV) system leads to a rapid
advance in the network industry. As practical applications,
Network Security Functions (NSFs), such as firewall, Intrusion
Detection System (IDS)/Intrusion Prevention System (IPS), and attack mitigation,
can also be provided as Virtual Network Functions (VNF) in the NFV system.
By the efficient virtual technology, these VNFs might be automatically
provisioned and dynamically migrated based on real-time security requirements.
This document presents a YANG data model to implement security functions
based on NFV.
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 RFC 2119 RFC8174
.
This document uses the terminology described in .
This document follows the guidelines of ,
uses the common YANG types defined in , and
adopts the Network Management Datastore Architecture (NMDA). The meaning of
the symbols in tree diagrams is defined in .
A Policy object represents a mechanism to express a Security Policy by
Security Administrator (i.e., I2NSF User) using Consumer-Facing Interface
toward Security Controller; the policy would be enforced on an NSF.
shows the XML instance of the
Policy object. The Policy object SHALL have following information:
This field identifies the name of this object.Date when this object was created or last modified.This field contains a list of rules. If the rule does not have a user-defined precedence, then any conflict must be manually resolved.
A policy is a container of Rules. In order to express a Rule, a Rule must have
complete information such as where and when a policy needs to be applied.
This is done by defining a set of managed objects and relationship among them.
A Policy Rule may be related segmentation, threat mitigation or telemetry data
collection from an NSF in the network, which will be specified as the sub-model
of the policy model in the subsequent sections.
shows the XML instance of the Rule object.
The rule object SHALL have the following information:
This field identifies the name of this object.This field indicates the date when this object was created or last modified.This field includes the information to determine whether the Rule Condition can be evaluated or not. See details in Section 3.1.This field contains all the checking conditions to apply to the objective traffic. See details in Section 4.2.This field identifies the action taken when a rule is matched. There is always an implicit action to drop traffic if no rule is matched for a traffic type. See details in Section 4.3. This field contains the information about IPsec type.
There are two types such as IPsec-IKE and IPsec-IKEless
.This field contains the onwer of the rule. For example, the person who created it, and eligible for modifying it.
The Event Object contains information related to scheduling a Rule.
The Rule could be activated based on a time calendar or security event
including threat level changes.
shows the XML instance
of the Event object. Event object SHALL have following information:
This field identifies the name of this object.This field indicates the date when this object was created or last modified.This field identifies whether the event of triggering policy enforcement is "ADMIN-ENFORCED", "TIME-ENFORCED" or "EVENT-ENFORCED".This field contains a time calendar such as "BEGIN-TIME" and "END-TIME" for one time enforcement or recurring time calendar for periodic enforcement.
This object represents Conditions that Security Administrator wants to apply
the checking on the traffic in order to determine whether the set of
actions in the Rule can be executed or not. The Condition Sub-model
consists of 3 different types of three containers each representing
different cases, such as general firewall and DDoS-mitigation cases,
and a case when the condition is based on the payload strings of packets.
Each containers have source-target and destination-target to represent
the source and destination for each case.
shows the XML instance of
the Condition object.
The Condition Sub-model SHALL have following information:
This field represents the general firewall case, where a security admin can set up firewall conditions using the information present in this field. The source and destination is represented as source-target and destination-target, each referring to the IP-address-based groups defined in the endpoint-group. This field represents the condition for DDoS mitigation, where a security admin can set up DDoS mitigation conditions using the information present in this field. The source and destination is represented as source-target and destination-target, each referring to the device-groups defined and registered in the endpoint-group. This field contains the payload string information. This information is useful when security rule condition is based on the string contents of incoming or outgoing packets. The source and destination is represented as source-target and destination-target, each referring to the payload-groups defined and registered in the endpoint-group.
This object represents actions that Security Admin wants to perform
based on certain traffic class.
shows the XML instance of the Action object. The Action object SHALL
have following information:
This field identifies the name of this object.This field indicates the date when this object was created or last modified.This field identifies the action when a rule is matched by an NSF. The action could be one of "PASS", "DROP", "ALERT", "MIRROR", and "LOG".
Multi-tenancy is an important aspect of any application that enables
multiple administrative domains in order to manage application resources.
An Enterprise organization may have multiple tenants or departments
such as Human Resources (HR), Finance, and Legal, with each tenant having
a need to manage their own Security Policies. In a Service Provider,
a tenant could represent a Customer that wants to manage its own Security
Policies. There are multiple managed objects that constitute
multi-tenancy aspects as shown in .
This section lists these objects and the relationship among these objects.
Below diagram shows an example of multi-tenancy in an Enterprise domain.
This object defines a boundary for the purpose of policy management
within a Security Controller. This may vary based on how the Security
Controller is deployed and hosted. For example, if an Enterprise hosts
a Security Controller in their network; the domain in this case could
just be the one that represents that Enterprise. But if a Cloud Service
Provider hosts managed services, then a domain could represent a single
customer of that Provider. shows
the XML instance of the Policy-Domain object. Multi-tenancy model should
be able to work in all such environments. The Policy-Domain object SHALL
have the following information:
Name of the organization or customer representing this domain. Address of the organization or customer. Contact information of the organization or customer. Date when this account was created or last modified. Authentication method to be used for this domain. It should be a reference to a "Policy-Management-Authentication-Method" object.
This object defines an entity within an organization.
The entity could be a department or business unit within an Enterprise
organization that would like to manage its own Policies due to
regulatory compliance or business reasons.
shows the XML instance of
the Policy-Tenant object. The Policy-Tenant object SHALL have
the following information:
Name of the Department or Division within an organization. Date when this account was created or last modified. This field identifies the domain to which this tenant belongs. This should be a reference to a Policy-Domain object.
This object defines a set of permissions assigned to a user in
an organization that wants to manage its own Security Policies.
It provides a convenient way to assign policy users to a job function
or a set of permissions within the organization.
shows the XML instance of
the Policy-Role object. The Policy-Role object SHALL have the
following information:
This field identifies the name of the role. Date when this role was created or last modified. This field identifies the access profile for the role. The profile grants or denies the permissions to access Endpoint Groups for the purpose of policy management or may restrict certain operations related to policy managements. There are two permission types, read-only and read-and-write, to choose from for each access-profile.
This object represents a unique identity of a user within an organization.
The identity authenticates with Security Controller using credentials
such as a password or token in order to perform policy management.
A user may be an individual, system, or application requiring access
to Security Controller. shows the XML
instance of the Policy-User object. The Policy-User object SHALL have
the following information:
Name of a user. Date when this user was created or last modified. User password for basic authentication. E-mail address of the user. This field identifies whether the user has domain-wide or tenant-wide privileges. This field should be a reference to a Policy-Role object that defines the specific permissions.
This object represents authentication schemes supported by
Security Controller.
shows the XML instance of the Policy Management Authentication
Method onject. This Policy-Management-Authentication-Method object
SHALL have the following information:
This field identifies name of this object. Date when this object was created or last modified. This field identifies the authentication methods. It could be a password-based, token-based, certificate-based or single sign-on authentication. This field indicates whether mutual authentication is mandatory or not. This field stores the information about server that validates the token submitted as credentials. This field stores the information about server that validates certificates submitted as credentials. This field contains the information about IPsec type. There are two types; 1) IPsec-IKE and IPsec-IKEless. This field stores the information about server that validates user credentials.
The Policy Endpoint Group is a very important part of building
User-Construct based policies. A Security Administrator would create
and use these objects to represent a logical entity in their business
environment, where a Security Policy is to be applied.
There are multiple managed objects that constitute a Policy's Endpoint Group
as shown in .
shows the XML instance of the
Endpoint-Group object.
This section lists these objects and relationship among them.
This object represents a User-Group. shows the XML instance of the User-Group object.
The User-Group object SHALL have the following information:
This field identifies the name of this object. Date when this object was created or last modified. This field identifies the IP address of a user. This field is a range of IP addresses of users.
This object represents a Device-Group. shows the XML instance of the Device-group object.The Device-Group object SHALL have the following information:
This field identifies the name of this object. Date when this object was created or last modified. This field identifies the IP address of a device. This field is a range of IP addresses of devices.
This object represents a location group based on either tag or other information.
shows the XML instance of
the Location-Group object.
The Location-Group object SHALL have the following information:
This field identifies the name of this object. Date when this object was created or last modified. to identify which continent the location group member is based at.
The threat prevention plays an important part in the overall security posture by reducing the attack surfaces. This information could come from various threat feeds (i.e., sources for obtaining the threat information), such as EmergingThreats.com or AlienVault.com. There are multiple managed objects that constitute this category. This section lists these objects and relationship among them.
shows the XML instance of a Threat-Prevention object.
This object represents a threat feed which provides signatures of
malicious activities. shows the XML instance of a Threat-feed-list.
The Threat-Feed object SHALL have the following information:
This field identifies the name of this object. Date when this object was created or last modified. This field identifies the information about the feed provider, it may be an external service or local server. This field identifies the information about the file types identified and reported by the threat-feed. This field contains the signatures of malicious programs or activities provided by the threat-feed.
This object represents a custom list created for the purpose of
defining exception to threat feeds. shows the XML instance of a Payload-content list.
The Payload-Content object SHALL have the following information:
This field identifies the name of this object. Date when this object was created or last modified. This field contains contents such as IP addresses or URL names.
Role-Based Access Control (RBAC) provides a powerful and centralized control
within a network. It is a policy neutral access control mechanism defined
around roles and privileges. The components of RBAC, such as role-permissions,
user-role and role-role relationships, make it simple to perform user assignments.
As shown in , a role represents a collection of permissions (e.g., accessing a file server or other particular resources). A role may be assigned to one or multiple users. Both roles and permissions can be organized in a hirarchy. A role may consists of other roles and permissions.
Following are the steps required to build RBAC:
Defining roles and permissions. Establishing relations among roles and permissions. Defining users. Associating rules with roles and permissions. assigning roles to users.
The main objective of this data model is to provide both
an information model and the corresponding YANG data model
of I2NSF Consumer-Facing Interface. This interface can be used
to deliver control and management messages between an I2NSF User
and Security Controller for the I2NSF User's high-level security
policies.
The semantics of the data model must be aligned with the information model of the Consumer-Facing Interface. The transformation of the information model was performed so that this YANG data model can facilitate the efficient delivery of the control or management messages.
This data model is designed to support the I2NSF framework that can be extended according to the security needs. In other words, the model design is independent of the content and meaning of specific policies as well as the implementation approach. This document suggests a VoIP/VoLTE security service as a use case for policy rule generation.
This section describes a YANG data model for Consumer-Facing Interface,
based on the information model of Consumer-Facing Interface to
Security Controller.
This section describes the XML instances for different policies examples that are delivered through Consumer-Facing Interface. The considered use cases are: VoIP/VoLTE security service, DDoS-attack mitigation, time-based firewall as a web-filter.
In order to create a rule of a security policy, it is essential to first register data (those which are used to form such rule) to the database. For example, The endpoint group consists of three different groups: user-group, device-group, and payload-group. Each of these groups have separate group members with information other than meta ("name" or "date"), such as ip-addresses or protocols used by devices. shows an example XML representation of the registered information for the user-group and device-group.
The first example scenario is to "block SNS access during business hours" using a time-based firewall policy. In this scenario, all users registered as "employee" in the user-group list are unable to access Social Networking Services (SNS) during the office hours. The XML instance is described below:
Time-based-condition Firewall
The policy name is "security_policy_for_blocking_sns".
The rule name is "block_access_to_sns_during_office_hours".
The Source-target is "employees".
The destination target is "sns-websites". "sns-websites" is the key which represents the list containing the information, such as URL, about sns-websites.
The action required is to "drop" any attempt to connect to websites related to Social networking.
The IPsec-method is set to "ikeless".
The second example scenario is to "block malicious VoIP/VoLTE packets coming to a company" using a VoIP policy. In this scenario, the calls comming from from VOIP and/or VOLTE sources with VOLTE IDs that are classified as malicious are dropped. The IP addresses of the employees and malicious VOIP IDs should be blocked are stored in the database or datastore of the enterprise. Here and the rest of the cases assume that the security administrators or someone responsible for the existing and newly generated policies, are not aware of which and/or how many NSFs are needed to meet the security requirements. represents the XML document generated from YANG discussed in previous sections. Once a high-level seucurity policy is created by a security admin, it is delivered by the Consumer-Facing Interface, through RESTCONF server, to the security controller. The XML instance is described below:
Custom-condition Firewall
The policy name is "security_policy_for_blocking_malicious_voip_packets".
The rule name is "Block_malicious_voip_and_volte_packets".
The Source-target is "malicious-id". This can be a single ID or a list of IDs, depending on how the ID are stored in the database. The "malicious-id" is the key so that the security admin can read every stored malicious VOIP IDs that are named as "malicious-id".
The destination target is "employees". "employees" is the key which represents the list containing information about employees, such as IP addresses.
The action required is "drop" when any incoming packets are from "malicious-id".
The IPsec-method is set to "ikeless".
The third example scenario is to "Mitigate HTTP and HTTPS flood attacks on a company web server" using a DDoS-attack mitigation policy. Here, the time information is not set because the service provided by the network should be maintained at all times. If the packets sent by any sources are more than the set threshold, then the admin can set the percentage of the packets to be dropped to safely maintain the service. In this scenario, the source is set as "any" to block any sources which send abnormal amount of packets. The destination is set as "web_server01". Once the rule is set and delivered and enforced to the nsfs by the securiy controller, the NSFs will monitor the incoming packet amounts and the destination to act according to the rule set. The XML instance is described below:
DDoS-condition Firewall
The policy name is "security_policy_for_ddos_attacks".
The rule name is "100_packets_per_second".
The destination target is "webservers". "webservers" is the key which represents the list containing information, such as IP addresses and ports, about web-servers.
The rate limit exists to limit the incoming amount of packets per second. In this case the rate limit is "100" packets per second. This amount depends on the packet receiving capacity of the server devices.
The Source-target is all sources which send abnormal amount of packets.
The action required is to "drop" packet reception is more than 100 packets per second.
The IPsec-method is set to "ikeless".
The data model for the I2NSF Consumer-Facing Interface is based
on the I2NSF framework , so the same security
considerations with the I2NSF framework should be included
in this document.
The data model needs a secure communication channel to protect
the Consumer-Facing Interface between the I2NSF User and
Security Controller.
This document requests IANA to register the following URI in the
"IETF XML Registry" :
This document requests IANA to register the following YANG
module in the "YANG Module Names" registry .
Interface to Network Security Functions (I2NSF) TerminologyRequirements for Client-Facing Interface to Security ControllerInformation Model of NSFs CapabilitiesSoftware-Defined Networking (SDN)-based IPsec Flow Protection
The following changes have been made from draft-ietf-i2nsf-consumer-facing-interface-dm-03:
This version added an I2NSF IPsec field for configuration and state data
for IPsec management (i.e., IPsec method such as IKE and IKEless
) in the I2NSF framework.
This work was supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government(MSIP) (No.R-20160222-002755, Cloud based Security Intelligence Technology Development for the Customized Security Service Provisioning).
This document is made by the group effort of I2NSF working group. Many people actively contributed to this document, such as Mahdi F. Dachmehchi and Daeyoung Hyun. The authors sincerely appreciate their contributions. The following are co-authors of this document:
Hyoungshick Kim
Department of Software
2066 Seo-ro Jangan-gu
Suwon, Gyeonggi-do 16419
Republic of Korea
EMail: hyoung@skku.edu
Seungjin Lee
Department of Electrical and Computer Engineering
2066 Seo-ro Jangan-gu
Suwon, Gyeonggi-do 16419
Republic of Korea
EMail: jine33@skku.edu
Jinyong Tim Kim
Department of Electrical and Computer Engineering
2066 Seo-ro Jangan-gu
Suwon, Gyeonggi-do 16419
Republic of Korea
EMail: timkim@skku.edu
Anil Lohiya
Juniper Networks
1133 Innovation Way
Sunnyvale, CA 94089
US
EMail: alohiya@juniper.net
Dave Qi
Bloomberg
731 Lexington Avenue
New York, NY 10022
US
EMail: DQI@bloomberg.net
Nabil Bitar
Nokia
755 Ravendale Drive
Mountain View, CA 94043
US
EMail: nabil.bitar@nokia.com
Senad Palislamovic
Nokia
755 Ravendale Drive
Mountain View, CA 94043
US
EMail: senad.palislamovic@nokia.com
Liang Xia
Huawei
101 Software Avenue
Nanjing, Jiangsu 210012
China
EMail: Frank.Xialiang@huawei.com