Internet DRAFT - draft-huang-netmod-acl
draft-huang-netmod-acl
Network Working Group L. Huang
Internet-Draft A. Clemm
Intended status: Informational Cisco Systems
Expires: March 08, 2014 A. Bierman
YumaWorks
September 04, 2013
YANG Data Model for Stateless Packet Filter Configuration
draft-huang-netmod-acl-03.txt
Abstract
A Stateless Packet Filter (SPF) determines which packets are allowed
to transit a system according to a set of rules, applying special
actions to packets as necessary. This document defines a YANG data
model for the configuration of Stateless Packet Filters on a device.
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at http://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
This Internet-Draft will expire on March 08, 2014.
Copyright Notice
Copyright (c) 2013 IETF Trust and the persons identified as the
document authors. All rights reserved.
Huang, et al. Expires March 08, 2014 [Page 1]
Internet-Draft yang-spf September 2013
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
This document may contain material from IETF Documents or IETF
Contributions published or made publicly available before November
10, 2008. The person(s) controlling the copyright in some of this
material may not have granted the IETF Trust the right to allow
modifications of such material outside the IETF Standards Process.
Without obtaining an adequate license from the person(s) controlling
the copyright in such materials, this document may not be modified
outside the IETF Standards Process, and derivative works of it may
not be created outside the IETF Standards Process, except to format
it for publication as an RFC or to translate it into languages other
than English.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Definitions and Acronyms . . . . . . . . . . . . . . . . . . 4
3. The Design of the Stateless Packet Filter Data Model . . . . 5
3.1. Overall Model Structure . . . . . . . . . . . . . . . . . 5
3.2. Data hierarchy . . . . . . . . . . . . . . . . . . . . . 6
3.3. Other Considerations . . . . . . . . . . . . . . . . . . 9
3.3.1. Extensibility . . . . . . . . . . . . . . . . . . . . 9
3.3.2. SPF Chain Support . . . . . . . . . . . . . . . . . . 9
3.3.3. SPF Test Extensions . . . . . . . . . . . . . . . . . 10
3.3.4. Attaching SPFs to interfaces . . . . . . . . . . . . 11
4. stateless-pf Module . . . . . . . . . . . . . . . . . . . . . 11
4.1. Features . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2. Types . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.3. Groupings . . . . . . . . . . . . . . . . . . . . . . . . 13
4.4. Containers . . . . . . . . . . . . . . . . . . . . . . . 13
4.4.1. spfs Container . . . . . . . . . . . . . . . . . . . 13
4.4.2. port-groups Container . . . . . . . . . . . . . . . . 14
4.4.3. timerange-groups Container . . . . . . . . . . . . . 14
4.4.4. ip-address-groups Container . . . . . . . . . . . . . 15
5. spf-ip module . . . . . . . . . . . . . . . . . . . . . . . . 16
5.1. Groupings . . . . . . . . . . . . . . . . . . . . . . . . 16
5.1.1. IP-SOURCE-NETWORK grouping . . . . . . . . . . . . . 16
5.1.2. IP-DESTINATION-NETWORK grouping . . . . . . . . . . . 17
5.1.3. DSCP-OR-TOS Grouping . . . . . . . . . . . . . . . . 17
Huang, et al. Expires March 08, 2014 [Page 2]
Internet-Draft yang-spf September 2013
5.1.4. IP-PFE-FILTERS Grouping . . . . . . . . . . . . . . . 18
5.2. augment . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.1. global-fragments leaf . . . . . . . . . . . . . . . . 21
6. spf-mac module . . . . . . . . . . . . . . . . . . . . . . . 23
6.1. MAC-SOURCE-NETWORK grouping . . . . . . . . . . . . . . . 23
6.2. MAC-DESTINATION-NETWORK grouping . . . . . . . . . . . . 24
6.3. augment . . . . . . . . . . . . . . . . . . . . . . . . . 25
7. spf-arp module . . . . . . . . . . . . . . . . . . . . . . . 25
7.1. augment . . . . . . . . . . . . . . . . . . . . . . . . . 25
8. Data Model Structure . . . . . . . . . . . . . . . . . . . . 25
9. SPF Examples . . . . . . . . . . . . . . . . . . . . . . . . 33
9.1. Configuration Example . . . . . . . . . . . . . . . . . . 33
10. Stateless-PF YANG Module . . . . . . . . . . . . . . . . . . 35
11. SPF-IP YANG Module . . . . . . . . . . . . . . . . . . . . . 48
12. SPF-MAC Configuration YANG Module . . . . . . . . . . . . . . 62
13. SPF-ARP Configuration YANG Module . . . . . . . . . . . . . . 68
14. COMMON-TYPES YANG Module . . . . . . . . . . . . . . . . . . 71
15. Security Considerations . . . . . . . . . . . . . . . . . . . 79
16. Open items from the previous revision . . . . . . . . . . . . 79
17. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 80
18. References . . . . . . . . . . . . . . . . . . . . . . . . . 80
18.1. Normative References . . . . . . . . . . . . . . . . . . 80
18.2. Informative References . . . . . . . . . . . . . . . . . 80
1. Introduction
This document defines a YANG [RFC6020] data model for the
configuration of Stateless Packet Filters (SPF).
A Stateless Packet Filter is a function that filters traffic on a
network device according to an ordered set of rules that define which
packets are to be permitted and which are to be denied. Each rule is
represented by a Packet Filter Entry (PFE). The sets of rules are
sometimes also referred to as "Access Control Lists" (ACL), the rules
as "Access Control Entries" (ACE) or simply "firewall rules". For
the purposes of this document, we will use the terms SPF, stateless-
pf and ACL interchangeably, as well as the terms PFE and ACE.
A PFE consists of two parts:
o A set of filters with a set of matching criteria that a packet
must satisfy for the rule to be applied.
o A set of actions (most commonly, a single action) that specifies
what to do with the packet when the matching criteria is met, for
example, to drop the packet.
Huang, et al. Expires March 08, 2014 [Page 3]
Internet-Draft yang-spf September 2013
There are different types of SPF, depending on which types of packets
they filter. Three of the most common types are covered in this
specification: MAC SPF, IP SPF, and ARP SPF.
o MAC SPFs: MAC SPFs are used to filter traffic using the
information in the Layer 2 header of each packet. MAC SPFs are by
default only applied to non-IP traffic; however, Layer 2
interfaces can be configured to apply MAC SPFs to all traffic.
o IP SPFs: IP SPFs are ordered sets of rules that can use to filter
traffic based on IP information in the Layer 3 header of packets.
The device applies IP SPFs only to IP traffic. IP SPF can be IPv4
or IPv6.
o ARP SPFs: ARP SPFs are used to filter Address Resolution Protocol
(ARP) traffic.
Not every device implements every type of SPF. The model for each
SPF type is therefore specified in its own YANG module. A device
will implement only the modules for the SPF types that it supports.
In addition, device implementations may vary greatly in terms of the
filter constructs that they support for any given SPF type.
Therefore, SPF YANG Module makes extensive use of the "feature"
construct which allows implementations to support those SPF
configuration features that lie within their capabilities.
The model can accommodate other SPF types beyond the ones that are
defined in this document. For this purpose, new SPF types can be
defined in their own modules which extend and augment the generic
portion of the model according to the same design pattern. This way,
the model serves as a framework that can be applied for any type of
Stateless Packet Filter.
2. Definitions and Acronyms
AFI: Address Field Identifier
ARP: Address Resolution Protocol
CoS: Class of Service
DSCP: Differentiated Services Code Point
ICMP: Internet Control Message Protocol
IGMP: Internet Group Management Protocol
IP: Internet Protocol
Huang, et al. Expires March 08, 2014 [Page 4]
Internet-Draft yang-spf September 2013
IPv4: Internet Protocol version 4
IPv6: Internet Protocol version 6
MAC: Media Access Control
PFE: Packet Filter Entry
QoS: Quality of Service
SPF: Stateless Packet Filter
TCP: Transmission Control Protocol
ToS: Type of Service
TTL: Time To Live
UDP: User Datagram Protocol
VLAN: Virtual Local Area Network
VRF: Virtual Routing and Forwarding
3. The Design of the Stateless Packet Filter Data Model
3.1. Overall Model Structure
The stateless-pf data model consists of five YANG modules. The first
module, "stateless-pf", defines generic SPF aspects which are common
to all SPFs regardless of their type, as well as a set of auxiliary
definitions. In effect, the module can be viewed as providing a
generic SPF "superclass".
Three other modules, "spf-ip", "spf-mac", and "spf-arp" , augment the
"stateless-pf" module with definitions that are specific to different
types of SPFs, specifically, SPFs for IP, MAC, and ARP, respectively.
These specifics are for the largest part reflected in the Packet
Filter Entries, that is, the rules which specify the filter criteria
that a packet must meet for the rule to be applied, and the actions
that are to be taken in case the filter matches. Keeping the modules
separate provides for a more modular data model than would be the
case if all types were combined into a single monolithic module.
To extend the model with other SPF types, additional modules that
augment the "stateless-pf" module can be defined, thus reflecting the
same model structure and following the same design pattern.
Huang, et al. Expires March 08, 2014 [Page 5]
Internet-Draft yang-spf September 2013
Finally, module "common-types" defines types that are used in the
stateless-pf data model but are not really specific to SPFs. These
definitions could potentially be of interest to other models as well;
keeping them in a separate module allows to import these definitions
independent of the support for SPFs.
3.2. Data hierarchy
The data hierarchy that is defined by the spf module is depicted in
the following Figure "SPF Model Structure", where brackets enclose
list keys, "rw" means configuration, "ro" means operational state
data, and "?" means optional node. Parentheses enclose choice and
case nodes. The structure is a collapsed structure and does not
depict all definitions; it is intended to illustrate the overall
structure. A fully expanded structure can be found in Data Model
Structure Section (Section 8).
module: stateless-pf
+--rw spfs
+--rw spf [name]
| +--rw name
| +--rw spf-type
| +--rw enable-capture-global?
| +--rw capture-session-id-global?
| +--rw (enable-match-counter-choices)?
| +--ro match?
|
|
+--rw port-groups
| +--rw port-group [name]
| +--rw name
| +--rw port-group-entry
+--rw timerange-groups
| +--rw timerange-group [name]
| +--rw name
| +--rw time-range
+--rw ip-address-groups
| +--rw ip-address-group [name]
| +--rw name
| +--rw afi?
| +--rw ip-address
SPF Model Structure
Data nodes in the stateless-spf module are contained under a single
container node, "spfs". This node contains a list, "spf". Each SPF
is represented by an element in that list and identified by a name
Huang, et al. Expires March 08, 2014 [Page 6]
Internet-Draft yang-spf September 2013
that serves as key to the list. Interfaces (which are not part of
the model, but for example defined per [if-config]) to which an SPF
is applied can then refer to the SPF using that name, respectively a
data type "spf-ref" introduced for that purpose. Each spf list
element has furthermore a type, as indicated through "spf-type". The
spf-type determines which types of PFEs can be can be contained in an
SPF. The PFE definitions themselves are provided by the spf-ip, spf-
mac, and spf-arp modules, which augment the spf definition in the spf
module accordingly. The subsequent data nodes in the spf list allow
to configure whether packets that match an SPF should be captured for
further analysis. Finally, the list contains an object that
maintains a counter of the number of SPF matches.
Auxiliary objects "port-groups", "ip-address-groups", "timerange-
groups" are used to define groupings of ports and of IP-addresses as
well as schedule information, respectively. They are in effect
convenience objects which allow PFEs to refer to groupings and
schedules by name, rather than needing to re-specify them in each PFE
where they apply.
The following figure depicts how different types of PFEs are inserted
into that structure. As indicated earlier, the corresponding
definitions are provided in separate modules that augment the spf
module. In the data structure, the augmenting module is indicated by
the prefix of the corresponding data nodes: "spf-ip", "spf-mac", and
"spf-arp", respectively. PFEs for IPv4 and for IPv6 are both defined
in the same module, spf-ip. While it would have been possible to
define each in its own separate module, it was a design decision to
combine them, as they share enough commonality that a separation
would have resulted in a considerable amount of definition
redundancy.
The figure does not depict objects not pertinent to that structure,
such as objects intended to make the definition of port groups
("port-groups"), timeranges ("time-range-groups"), and IP address
groups ("ip-address-groups") reusable, as well as objects that are
contained in spf list elements, such as "name" and "enable-capture-
global".
module: stateless-pf
+--rw spfs
+--rw spf [name]
| +--rw spf-ip:afi
| +--rw spf-ip:ipv6-pfes
| | +--rw spf-ip:ipv6-pfe [name]
| | +--rw spf-ip:name
| | +--rw (remark-or-ipv6-case)?
| | +--:(remark)
Huang, et al. Expires March 08, 2014 [Page 7]
Internet-Draft yang-spf September 2013
| | | +--rw spf-ip:remark
| | +--:(ipv6-pfe)
| | | +--rw spf-ip:filters
| | | +-- filter parameters
| | | +--rw spf-ip:actions
| | | +-- action parameters
| | +-- ro spf-ip:match
module: stateless-pf
+--rw spfs
+--rw spf [name]
| +--rw spf-ip:afi
| +--rw spf-ip:ipv4-pfes
| | +--rw spf-ip:ipv4-pfe [name]
| | +--rw spf-ip:name
| | +--rw (remark-or-ipv4-pfe)?
| | +--:(remark)
| | | +--rw spf-ip:remark
| | +--:(ipv4-pfe)
| | | +--rw spf-ip:filters
| | | +-- filter parameters
| | | +--rw spf-ip:actions
| | | +-- action parameters
| | +-- ro spf-ip:match
module: stateless-pf
+--rw spfs
+--rw spf [name]
| +--rw spf-mac:mac-pfes
| | +--rw spf-mac:mac-pfe [name]
| | +--rw spf-mac:name
| | +--rw (remark-or-mac-pfe)?
| | +--:(remark)
| | | +--rw spf-mac:remark
| | +--:(mac-pfe)
| | | +--rw spf-mac:filters
| | | +-- filter parameters
| | | +--rw spf-mac:actions
| | | +-- action parameters
| | +-- ro spf-mac:match
module: stateless-pf
+--rw spfs
+--rw spf [name]
| +--rw spf-arp:arp-pfes
| | +--rw spf-arp:arp-pfe [name]
Huang, et al. Expires March 08, 2014 [Page 8]
Internet-Draft yang-spf September 2013
| | +--rw spf-arp:name
| | +--rw (remark-or-arp-pfe)?
| | +--:(remark)
| | | +--rw spf-arp:remark
| | +--:(arp-pfe)
| | | +--rw spf-arp:filters
| | | +-- filter parameters
| | | +--rw spf-arp:actions
| | | +-- action parameters
| | +-- ro spf-arp:match
Model structure - different SPF types
As is evident from Figure "Model structure - different SPF types",
the same generic design pattern is reflected in every SPF type. Each
SPF contains a list of PFEs, identified by a name by which PFEs in
the list are ordered. Each PFE consists either of a remark or of an
actual access control rule. Remarks are in effect comment lines
inside an SPF that are intended for human or administrator
consumption. They are included in the YANG module to maintain
consistency with CLI. Access control rules, on the other hand,
consist of a left hand side ("filters") that specifies a set of
matching criteria and a right hand side ("actions") that specifies
the action to take when matching criteria are met. An overview of
the full list of filter and parameters is given in Section 8.
Since the design pattern for each SPF type is the same, an
alternative design to the YANG modules would have been to extend the
"spf" module to include the data nodes up to the level depicted in
Figure "Model structure - different SPF types", as the real
distinction occurs in the filter and action parameters that occur
below it. In that case, however, the corresponding data nodes would
have had to contend with more complex conditions. The modules
defined here aim at keeping complexity of definitions within the
modules as low as possible, at the price of repeating a few data
nodes that provide the overall top level structure.
3.3. Other Considerations
3.3.1. Extensibility
If needed, the model can be extended for other types of SPFs in
straightforward manner. New types of SPFs can be defined in
additional YANG modules that apply the same design patterns much in
the same way as in the case of IP, MAC, and ARP SPFs.
3.3.2. SPF Chain Support
Huang, et al. Expires March 08, 2014 [Page 9]
Internet-Draft yang-spf September 2013
SPF chains are used in some application domains. SPF chains are not
included in the data model, but could be accommodated in the model
through extensions in a straightforward way.
SPF chains work roughly as follows. In an SPF chain, as an
alternative to an action, an PFE can point to another SPF. If a
packet matches the filter condition, it is subjected to the other
SPF. If the other SPF contains an PFE that matches, that action is
executed. If there is no match, processing is returned to the first
SPF and processing continues with the subsequent PFEs until a match
is found. This way, chained SPFs can be considered as a special form
of "SPF subroutine".
An example of an SPF chain might be a rule that contains a filter for
a specific destination port number in an IP packet, then invokes
another SPF that contains a specific set of firewall rules for
traffic directed at that particular port. Even though the data model
for SPF presented in this document uses a flat list of PFE in each
SPF, the actions in the model can be augmented to support SPF chains.
The model can be extended with SPF chains roughly as follows: A new
spf-chaining action is introduced, represented as a leaf whose value
contains a reference to an SPF as a parameter. Below is an example
of how the spf-ip model could be extended to support SPF chains for
ip-v4:
augment "/spf:spfs/spf:spf/spf-ip:ipv4-pfes" +
"/spf-ip:ipv4-pfe/spf-ip:actions" {
leaf chain {
type spf-ref ;
description "Reference to another SPF name to chain the PFEs";
}
}
For SPFs that are expected to not terminate when no PFE matches, but
return processing to the invoking SPF, an optional SPF parameter can
be introduced that indicates for chained SPFs which chaining behavior
should apply.
3.3.3. SPF Test Extensions
Given the complexity of SPFs in many deployments, debugging SPFs and
assessing whether an SPF has the actual desired effect can be a
challenge. In order to facilitate those tasks and allow to check
whether an SPF has indeed the intended effect, an additional
Huang, et al. Expires March 08, 2014 [Page 10]
Internet-Draft yang-spf September 2013
administrative function that allows applications and users to test a
packet against the SPF can be introduced. The function can take the
form of an RPC which takes as input parameter a leaf with the
reference to the SPF that is to be tested, and a leaf with a packet.
The output parameter includes a leaf indicating the action that is
taken as a result, as well as a leaf with the reference to the
matching PFE.
3.3.4. Attaching SPFs to interfaces
SPFs typically do not exist in isolation. Intead, they are
associated with a certain scope in which they are applied, for
example, an interface of a set of interfaces. How to attach an SPF
to an interface (or other system artifact) is outside the scope of
this model, as it depends on the specifics of the system model that
is being applied. However, in general, the general design pattern
will involve adding a data node with a reference, or set of
references, to SPFs that are to be applied to the interface. For
this purpose, the type definition "spf-ref" can be used.
For example, to attach an SPF to an interface as defined per the data
model [if-config], the following steps can be applied:
o Introduce a new YANG module to extend the interface configuration
YANG module.
o Import modules "interfaces" [if-config] (prefix: "if") and
"stateless-pf" (prefix: "spf").
o Augment list "interface" (/if:interfaces/if:interface) with a
leaf-list of type "spf:spf-ref".
4. stateless-pf Module
Module "stateless-pf" is a top container module for all SPFs. It
contains a container "spfs" with a list "spf" of named SPFs. Modules
"spf-ip", "spf-mac", and "spf-arp" augment this list with the objects
that are specific to each respective type of SPF. In addition,
module "spf" also defines a set of features, reusable types, and
reusable groupings.
4.1. Features
When it comes to SPF implementations, a wide range of different
capabilities exists across devices. For example, not every device
implements every type of SPF. Some devices may support time-based
SPFs that are only in effect during specified times, others may not.
Huang, et al. Expires March 08, 2014 [Page 11]
Internet-Draft yang-spf September 2013
In order to accommodate this wide range of capabilities, this data
model makes extensive use of the "feature" construct. The defined
features allow implementations to declare which capabilities they
support, and only support the corresponding portions of the data
model.
4.2. Types
The definition of SPFs requires a number of new data types introduced
in this data model. Table 1 depicts data types that are unique to
SPFs. Table 2 depicts data types that are required by SPFs, but not
specific to them, and that may hence be reused by other models.
Those data types are defined in module "common-types". For details
of each type, please refer to the corresponding typedef descriptions
and references in the model.
+----------------------+------------------------------+
| YANG type | base type |
+---------------------+-------------------------------+
| spf-comparator | enumeration |
| spf-action | enumeration |
| spf-remark | string |
| spf-type-ref | identityref |
| spf-ref | leafref |
| port-group-ref | leafref |
| ip-address-group-ref | leafref |
| time-range-Ref | leafref |
| weekdays | bits |
| spf-name-string | string |
+--------------------- +------------------------------+
Table 1
+----------------------+------------------------------+
| YANG type | base type |
+---------------------+-------------------------------+
| cos | uint8 |
| tos | uint8 |
| precedence | uint8 |
| tcp-flag-type | enumeration |
| ether-type | string |
| ip-protocol | uint8 |
| igmp-code | uint8 |
| icmp-type | uint32 |
| icmp-code | uint32 |
| vlan-identifier | uint16 |
| time-to-live | uint32 |
Huang, et al. Expires March 08, 2014 [Page 12]
Internet-Draft yang-spf September 2013
+--------------------- +------------------------------+
Table 2
4.3. Groupings
The data model defines two groupings, PFE-COMMON and FILTER-COMMON.
o PFE-COMMON is a collection of nodes that should be added to every
PFE list entry. PFE-COMMON contains the actions container and a
read-only match leaf. The actions container contains two leaves.
* An "action" leaf that specifies what to do with the packet when
the matching criteria is met, for example, to drop the packet.
* A "log" leaf that indicates whether to create a log entry when
an pfe filter matches. (Some devices may not support a log
capability. Hence support of this leaf is conditional on
declaration of a corresponding feature, as indicated by use of
the "if-feature" construct.)
o FILTER-COMMON is a collection of nodes that should be added to
every 'filters' container within each PFE list entry.
4.4. Containers
4.4.1. spfs Container
Container "spfs" contains a list "spf" of named SPFs. Each list
eleement "spf" contains the following global leaves. The list
elements are augmented with additional data nodes defined in modules
"spf-arp", "spf-mac", and "spf-ip".
o name
o spf-type
o enable-capture-global
o capture-session-id-global
o enable-match-counter-choices: The difference of these two choices
is that "enable-match-counter" indicates to collect total match
statistics for all pfes, whereas "enable-per-entry-match-counter"
indicates to collect match statistics for each PFE.
o match
Huang, et al. Expires March 08, 2014 [Page 13]
Internet-Draft yang-spf September 2013
4.4.2. port-groups Container
Container "port-groups" allows to classifying protocol port into
groups. It contains a sequence of "port-group" data nodes. Each
"port-group" defines a range of ports and can be referred to by name.
Multiple PFEs can refer to the same port group. The following is a
Netconf XML example of port-groups and how it is referred to from an
PFE.
<src-port-group-name>
<port-group-name>port-tunnel1</port-group>
</src-port-group-name>
<port-groups>
<port-group>
<name>port-tunnel1</name>
<port-group-entry>
<name>http-proxy</name>
<port-lower>21</port-lower>
<port-upper> 22</port-upper>
</port-group-entry>
</port-group>
</port-groups>
4.4.3. timerange-groups Container
Container "timerange-groups" container contains a list, "timerange-
group". Eeach of its elements defines a sequence of time ranges,
"time-range". Each time-range object consists of either a remark
(comments for the time range), or of an absolute time for start or
end (or both) of the time range, or a periodic time for start or end
or both. Object "remark" contains administrator-provided comments
for the time-range that will be kept in the device. Like with port
groups, the same time-range can be reused by different PFEs. The
following is a Netconf XML example of a timerange group that contains
a remark and a single time range.
<timerange-groups>
<timerange-group>
<name>weekday</name>
<time-range>
<name>10</name>
<remark> email server maintenance</remark>
</time-range>
<time-range>
<name>20</name>
Huang, et al. Expires March 08, 2014 [Page 14]
Internet-Draft yang-spf September 2013
<periodic>
<weekday>
Monday Tuesday Wednesday Thursday Friday
</weekday>
<start> 21:00:00</start>
<end> 24:00:00</end>
</periodic>
</time-range>
</timerange-group>
</timerange-groups>
4.4.4. ip-address-groups Container
Container "ip-address-groups" contains is list "ip-address-group" of
named IP address groups. Each IP address group is a sequence of
pairs "ip-address" and "mask", or a pair of "host" and "host-
address". Each IP address group can be referred from an PFE by name.
The following is a Netconf XML example of an IP address group and how
it is referred to from an PFE.
<ip-address-groups>
<ip-address-group>
<name>Email-Server-IPV4</name>
<ip-addresses>
<ip-address>
<name>10</name>
<ip-address>128.107.0,0</ip-address>
<ip-mask>255.255.0.0</ip-mask>
</ip-address>
<ip-address>
<name>20</name>
<ip-address>139.207.0.0</ip-address>
<ip-mask>255.255.0.0</ip-mask>
</ip-address>
</ip-addresses>
</ip-address-group>
</ip-address-groups>
<ip-pfe>
<name>100</name>
<afi>ipv4</afi>
<actions>permit</actions>
<filters>
<ip-source-group>Email-Server-IPV4</ip-source-group>
<ip-dest-any/>
Huang, et al. Expires March 08, 2014 [Page 15]
Internet-Draft yang-spf September 2013
</filters>
</ip-pfe>
5. spf-ip module
spf-ip is the module that defines IP-SPF. It augments spf list in
spf module.
5.1. Groupings
5.1.1. IP-SOURCE-NETWORK grouping
IP-SOURCE-NETWORK
+--rw (source-address-host-group)?
+--:(source-ip)
| +--rw ip-source-address inet:ip-address
| +--rw ip-source-mask inet:ip-address
+--:(ip-source-any)
| +--rw ip-source-any empty
+--:(source-host)
| +--:(ip-src-host-address-or-name)
| +--:(ip-source-host-address)
| +--rw ip-source-host-address inet:ip-address
| +--:(ip-source-host-name)
| +--rw ip-source-host-name inet:domain-name
+--:(source-group)
+--rw ip-source-group? ip-address-group-ref
IP-SOURCE-NETWORK is a reusable grouping. It allows five ways to
specify a network: ip with mask, any network, host-name or host
address, reference to a predefined ip address group. Here are valid
example instances:
o ip with mask:
<ip-source-address>192.168.1.0</ip-source-address>
<ip-source-mask>255.255.255.0</ip-source-mask>
o any network:
<ip-source-any/>
o host-name:
Huang, et al. Expires March 08, 2014 [Page 16]
Internet-Draft yang-spf September 2013
<ip-source-host-name>switch1</ip-source-host-name>
o host-address:
<ip-source-host-address>192.168.1.2</ip-source-host-address>
o reference to a predefined ip address group (Email-Server-IPV4 is
defined in Section 4.4.4 ):
<ip-source-group>Email-Server-IPV4</ip-source-group>
5.1.2. IP-DESTINATION-NETWORK grouping
IP-DESTINATION-NETWORK
+--rw (dest-address-host-group)?
+--:(dest-ip)
| +--rw ip-dest-address inet:ip-address
| +--rw ip-dest-mask? inet:ip-address
+--:(ip-dest-any)
| +--rw ip-dest-any empty
+--:(dest-host)
| +--:(ip-dest-host-address-or-name)
| +--:(ip-dest-host-address)
| +--rw ip-dest-host-address inet:ip-address
| +--:(ip-dest-host-name)
| +--rw ip-dest-host-name inet:domain-name
+--:(group)
+--rw ip-dest-group? ip-address-group-ref
IP-DESTINATION-ADDRESS is a reusable grouping. Its structure is
similar to IP-SOURCE-NETWORK. The reason to have both IP-SOURCE-
NETWORK and IP-DESTINATION-NETWORK groupings is to allow "ip-source-
address" and "ip-destination-address" leaves to appear in the same
container. For example:
<filters>
<ip-source-address>192.168.1.0</ip-source-address>
<ip-source-mask>255.255.255.0</ip-source-mask>
<ip-dest-address>any</ip-dest-address>
</filters>
5.1.3. DSCP-OR-TOS Grouping
Huang, et al. Expires March 08, 2014 [Page 17]
Internet-Draft yang-spf September 2013
DSCP-OR-TOS grouping defines a choice, "dscp-or-tos". It allows two
ways to filter for a QoS packet:
o dscp: Match packet on DSCP value.
o tos: Match packet on TOS and precedence value.
The typedef for "tos" and "precedence" is defined in module "common-
types", which could be deprecated should IETF define a separate set
of definitions.
5.1.4. IP-PFE-FILTERS Grouping
IP-PFE-FILTERS
+--rw protocol? c-types:ip-protocol
+--spf:FILTER-COMMON
+--rw fragments? empty
+--rw time-range? spf:Time-Range-Ref
+-- (src-ports)?
| +--rw (port-number-or-range)?
| | +--:(port-number-range)
| | | +--rw src-port-lower? inet:port-number
| | | +--rw src-port-upper? inet:port-number
| +--:(port-number)
| | +--rw src-comparator comparator
| | +--rw src-port? inet:port-number
| +-- :(port-group-ref)
| +--src-port-group-name
+-- (des-ports)?
| +--rw (port-number-or-range)?
| | +--:(port-number-range)
| | | +--rw des-port-lower? inet:port-number
| | | +--rw des-port-upper? inet:port-number
| +--:(port-number)
| | +--rw des-comparator comparator
| | +--rw des-port? inet:port-number
| +-- :(by-name)
| +-- des-port-group-name
+--rw icmp-type? c-types:icmp-type
+--rw icmp-code? c-types:icmp-type
+--rw (packet-length-or-range)?
| +--:(length)
| | +--rw packet-length-comparator spf:Comparator
| | +--rw packet-length uint32
| +--:(range)
| +--rw packet-length-upper uint32
| +--rw packet-length-lower uint32
+--rw tcp-flag-value? c-types:tcp-flag-type
Huang, et al. Expires March 08, 2014 [Page 18]
Internet-Draft yang-spf September 2013
+--rw tcp-flag-mask? c-types:tcp-flag-type
+--rw tcp-flag-operation? enumeration
+--rw (ttl-value-or-range)?
+--:(value)
| +--rw ttl-comparator? spf:spf-comparator
| +--rw ttl-value? c-types:Time-to-Live
+--:(range)
+--rw ttl-value-lower? c-types:Time-to-Live
+--rw :ttl-value--upper? c-types:Time-to-Live
IP-PFE-FILTERS defines the following leaves that are used by both by
IPv4 and IPv6 PFEs:
o protocol
o spf:FILTER-COMMON: see Section 4.3
o fragments: When present, it matches the non-initial fragment.
o time-range: Enable packet capture on this filter for a timerange-
group by name. time-range is Time-Range-Ref type which is a
leafref.
o src-ports choice: Allows the following three ways to define a
group of ports.
* port-number-range: Use "src-port-lower" and "src-port-upper"
leaves to specify a port range. The value of "src-port-lower"
has to be less than or equal the value of "src-port-upper".
* port-number: Use "comparator" and "src-port" leaves to specify
a port range. See Comparator typedef in the model for the
possible values the "comparator" leaf.
* port range ref: Refer to a named port group that is defined
using port-groups. For example:
<port-group-name>port-tunnel1</port-group-name>
o dest-ports choice: Analogous to "src-ports".
o packet-length-or-range: Allows two ways to specify packet length
range.
Huang, et al. Expires March 08, 2014 [Page 19]
Internet-Draft yang-spf September 2013
case length: Use comparator and a single packet-length to
specify the range.
case range: Use packet-length-lower and packet-length-upper to
specify a range. The value of packet-length-lower must be
lower than or equal to the value of packet-length-upper.
o icmp-type
o icmp-code
o packet-length-or-range choice
o tcp-flag-value: tcp-flag-value, tcp-flag-mask and tcp-flag-
operation allow to match any combination of packet tcp flag
values.
The following example is to match the packet
tcp flag ack=1, syn=1, and fin=0;
<tcp-flag-value> ack syn <tcp-flag-value>
<tcp-flag-mask>ack syn fin</tcp-flag-mask>
<tcp-flag-operation>match-all</tcp-flag-operation>
o tcp-flag-mask
o tcp-flag-operation
o ttl-value-or-range
5.2. augment
The module "spf-ip" augments the definition of data node "/spf:spfs/
spf:spf" with additional leaves and subcomponents.
o afi
o ipv6-pfes: It contains a list of ipv6-pfe. Each ipv6-pfe is
either a remark or a real access control filters. The case
ipv6-pfe defines the filters and actions for ipv6-pfe. The pfe
uses filters defined in grouping IP-SOURCE-NETWORK, IP-
DESTINATION-NETWORK, IP-PFE-FILTERS, DSCP-OR-TOS. In addition, it
also allows filter on igmp-type and flow-label,
o ipv4-pfes: ipv4-pfe has similar structure to ipv6-pfes.
o global-fragments
Huang, et al. Expires March 08, 2014 [Page 20]
Internet-Draft yang-spf September 2013
5.2.1. global-fragments leaf
global-fragments is an optional leaf. It has an enumeration value of
not-set, permit-all, deny-all. not-set is the default value. When
the global-fragments is permit-all or deny-all, it is to permit or
deny the implicit pfe fragment filter. Here is an example of
implicit pfe and how the implicit pfe is affected when global-
fragments is set.
Example 1: The spf configuration from the management interface with
global-fragments is absent.
YANG instance of this cli configuration:
<spfs>
<spf>
<name>fragment_test1</name>
<afi>ipv4</afi>
<spf-type>ip-spf</spf-type>
<ip-pfes>
<name>10</name>
<actions>
<action>permit</action>
</actions>
<filters>
<ip-source-address>192.168.5.0</ip-source-address>
<ip-source-mask>255.255.255.0</ip-source-mask>
<ip-dest-address>any</ip-dest-address>
</filters>
</ip-pfes>
<ip-pfes>
<name>20</name>
<actions>
<action>permit</action>
</actions>
<filters>
<ip-source-address>189.168.0.0</ip-source-address>
<ip-source-mask>255.255.0.0</ip-source-mask>
<ip-dest-address>any</ip-dest-address>
<fragments/>
</filters>
</ip-pfes>
</spf>
</spfs>
By taking all the tags out, the above yang can be express in a
summary of cli format like the following:
Huang, et al. Expires March 08, 2014 [Page 21]
Internet-Draft yang-spf September 2013
fragment_test1 ip-spf ipv4
10 permit ip 192.168.5.0 255.255.255.0 any
20 permit ip 189.168.0.0 255.255.0.0 any fragment.
The spf configuration together with implicit pfe in the device will
be:
fragment_test1 ip-spf ipv4
10 permit ip 192.168.5.0 255.255.255.0 any
11 permit ip 192.168.5.0 255.255.255.0 any fragment
20 permit ip189.168.0.0 255.255.0.0 any fragment.
100 deny any any
110 deny any any fragment
Notice three lines of configuration. 11, 100 and 110, are implicit.
Example 2: The spf configuration from the management interface with
global-fragments
<spfs>
<spf>
<name>fragment_test2</name>
<spf-type>ip-spf</spf-type>
<global-fragments>deny-all</global-fragments>
<afi>ipv4</afi>
<ip-pfes>
<name>10</name>
<actions>
<action>permit</action>
</actions>
<filters>
<ip-source-address>192.168.5.0</ip-source-address>
<ip-source-mask>255.255.255.0</ip-source-mask>
<ip-dest-address>any</ip-dest-address>
</filters>
</ip-pfes>
<ip-pfes>
<name>20</name>
<actions>
<action>permit</action>
</actions>
<filters>
Huang, et al. Expires March 08, 2014 [Page 22]
Internet-Draft yang-spf September 2013
<ip-source-address>189.168.0.0</ip-source-address>
<ip-source-mask>255.255.0.0</ip-source-mask>
<ip-dest-address>any</ip-dest-address>
<fragments/>
</filters>
</ip-pfes>
</spf>
</spfs>
The spf configuration in the device with implicit aces. The deny-all
void "11 permit ip 1.1.1.1/16 any fragment" pfe in previous example.
By taking all the tags out, the above yang can be express in a
summary of cli format like the following:
fragment_test2 ip-spf ipv4 deny-all
10 permit ip 192.168.5.0 255.255.255.0 any
20 permit ip 189.168.0.0 255.255.0.0 any fragment.
The spf configuration together with implicit pfe in the device will
be:
fragment_test2 ip-spf ipv4
10 permit ip 192.168.5.0 255.255.255.0 any
20 permit ip 189.168.0.0 255.255.0.0 any fragment.
100 deny any any
110 deny any any fragment
6. spf-mac module
6.1. MAC-SOURCE-NETWORK grouping
MAC-SOURCE-NETWORK
+--rw (source-network)?
+--:(source-mac)
| +--rw source-address yang:mac-address
| +--rw source-address-mask yang:mac-address
+--:(source-any)
| +--rw source-any empty
+--:(source-host)
+--rw spf-mac:source-host-name inet:host
MAC-SOURCE-ADDRESS is a reusable grouping. It allows to express the
three kinds network.
Huang, et al. Expires March 08, 2014 [Page 23]
Internet-Draft yang-spf September 2013
any network: use source-any to express any network.
<mac-source-kind>any</mac-source-kind>
single host network.
<source-host-name>my-host</source-host-name>
host address with a mask.
<source-address>0180.c200.000</source-address>
<source-address-mask>0000.0000.0000</source-address-mask>
6.2. MAC-DESTINATION-NETWORK grouping
MAC-DESTINATION-NETWORK
+--rw (dest-network)?
+--:(address)
| +--rw dest-address yang:mac-address
| +--rw dest-address-mask yang:mac-address
+--:(dest-any)
| +--rw dest-any empty
+--:(host)
+--rw spf-mac:dest-host-name inet:host
MAC-DESTINATION-ADDRESS is a reusable grouping similar to MAC-SOURCE-
ADDRESS. The reason to have both MAC-SOURCE-ADDRESS and MAC-
DESTINATION-ADDRESS grouping is to allow source-address and
destination-address leaves appear in the same container. For
example:
<filters>
<source-address>0180.c200.000</source-address>
<source-address-mask>0000.0000.0000</source-address-mask>
<dest-any/>
</filters>
Huang, et al. Expires March 08, 2014 [Page 24]
Internet-Draft yang-spf September 2013
6.3. augment
The module "spf-mac" augments the definition of data node "/spf:spfs/
spf:spf" with additional leaves and subcomponents. spf-mac has
similar structure as spf-ipv4 and spf-ipv6 except the filters are
different. mac-pfe has filters defined in grouping MAC-SOUCE-NETWORK,
MAC-DESTINATION-NETWORK, spf:FILTER-COMMON, ethertype-mask, cos,
time-range, and vlan.
7. spf-arp module
7.1. augment
The module "spf-arp" augments the definition of data node "/spf:spfs/
spf:spf" with additional leaves and subcomponents.
augment "/spf:spfs/spf:spf"
+--rw spf-arp:arp-pfes
+--rw spf-arp:arp-pfe [name]
+--rw spf-arp:name spf:spf-name-string
+--rw (remark-or-arp-pfe)?
+--:(remark)
| +--rw spf-arp:remark? spf:spf-remark
+--:(arp-pfe)
+--rw filters
| +--rw direction? enumeration
| +--spf-ip:IP-SOURCE-NETWORK
| +--spf-ip:IP-DESTINATION-NETWORK
| +--spf-mac:MAC-SOURCE-NETWORK
| +--spf-mac:MAC-DESTINATION-NETWORK
| +--spf:FILTER-COMMON
+spf:PFE-COMMON
8. Data Model Structure
The combined data model for SPF configuration is structured as
follows. "spf" defines the generic components of an spf system. "spf-
ip", "spf-mac", "spf-arp" augment the "spf" module with additional
data nodes that are needed for ip, mac, and arp spf respectively.
module: stateless-pf
+--rw spfs
+--rw spf [name]
| +--rw name
| +--rw spf-type
| +--rw enable-capture-global?
Huang, et al. Expires March 08, 2014 [Page 25]
Internet-Draft yang-spf September 2013
| +--rw capture-session-id-global?
| +--rw (enable-match-counter-choices)?
| | +--:(match)
| | | +--rw enable-match-counter?
| | +--:(per-entry-match)
| | +--rw enable-per-entry-match-counter?
| +--ro match?
| +--rw spf-ip:afi?
| +--rw spf-ip:ipv6-pfes
| | +--rw spf-ip:ipv6-pfe [name]
| | +--rw spf-ip:name spf:spf-name-string
| | +--rw (remark-or-ipv6-case)?
| | +--:(remark)
| | | +--rw spf-ip:remark? spf:spf-remark
| | +--:(ipv6-pfe)
| | +--rw spf-ip:filters
| | | +--rw (source-address-host-group)
| | | | +--:(source-ip)
| | | | | +--rw spf-ip:ip-source-address
| | | | | +--rw spf-ip:ip-source-mask
| | | | +--:(ip-source-any)
| | | | | +--rw spf-ip:ip-source-any?
| | | | +--:(source-host)
| | | | | +--rw (ip-src-address-or-name)
| | | | | +--:(ip-source-host-address)
| | | | | | +--rw spf-ip:ip-source-host-address?
| | | | | +--:(ip-source-host-name)
| | | | | +--rw spf-ip:ip-source-host-name?
| | | | +--:(source-group)
| | | | +--rw spf-ip:ip-source-group?
| | | +--rw (dest-address-host-group)
| | | | +--:(dest-ip)
| | | | | +--rw spf-ip:ip-dest-address
| | | | | +--rw spf-ip:ip-dest-mask
| | | | +--:(ip-dest-any)
| | | | | +--rw spf-ip:ip-dest-any?
| | | | +--:(dest-host)
| | | | | +--rw (ip-dest-address-or-name)
| | | | | +--:(ip-dest-host-address)
| | | | | | +--rw spf-ip:ip-dest-host-address?
| | | | | +--:(ip-dest-host-name)
| | | | | +--rw spf-ip:ip-dest-host-name?
| | | | +--:(dest-group)
| | | | +--rw spf-ip:ip-dest-group?
| | | +--rw spf-ip:protocol?
| | | +--rw spf-ip:enable-capture?
| | | +--rw spf-ip:capture-session-id?
| | | +--rw spf-ip:fragments?
Huang, et al. Expires March 08, 2014 [Page 26]
Internet-Draft yang-spf September 2013
| | | +--rw spf-ip:time-range?
| | | +--rw (src-ports)?
| | | | +--:(port-number-range)
| | | | | +--rw spf-ip:src-port-lower
| | | | | +--rw spf-ip:src-port-upper
| | | | +--:(port-number)
| | | | | +--rw spf-ip:src-comparator
| | | | | +--rw spf-ip:src-port
| | | | +--:(port-group-ref)
| | | | +--rw spf-ip:src-port-group-name
| | | +--rw (dest-ports)?
| | | | +--:(port-number-range)
| | | | | +--rw spf-ip:des-port-lower
| | | | | +--rw spf-ip:des-port-upper
| | | | +--:(port-number)
| | | | | +--rw spf-ip:des-comparator
| | | | | +--rw spf-ip:des-port
| | | | +--:(port-group-ref)
| | | | +--rw spf-ip:des-port-group-name
| | | +--rw spf-ip:icmp-type?
| | | +--rw spf-ip:icmp-code?
| | | +--rw (packet-length-or-range)?
| | | | +--:(length)
| | | | | +--rw spf-ip:packet-length-comparator
| | | | | +--rw spf-ip:packet-length
| | | | +--:(range)
| | | | +--rw spf-ip:packet-length-upper
| | | | +--rw spf-ip:packet-length-lower
| | | +--rw spf-ip:tcp-flag-value?
| | | +--rw spf-ip:tcp-flag-mask?
| | | +--rw spf-ip:tcp-flag-operation?
| | | +--rw (ttl-value-or-range)?
| | | | +--:(value)
| | | | | +--rw spf-ip:ttl-comparator?
| | | | | +--rw spf-ip:ttl-value?
| | | | +--:(range)
| | | | +--rw spf-ip:ttl-value-lower?
| | | | +--rw spf-ip:ttl-value--upper?
| | | +--rw (dscp-or-tos)?
| | | | +--:(dscp)
| | | | | +--rw spf-ip:dscp?
| | | | +--:(tos)
| | | | +--rw spf-ip:tos?
| | | | +--rw spf-ip:precedence?
| | | +--rw spf-ip:igmp-type?
| | | +--rw spf-ip:flow-label?
| | +--rw spf-ip:actions
| | | +--rw spf-ip:action
Huang, et al. Expires March 08, 2014 [Page 27]
Internet-Draft yang-spf September 2013
| | | +--rw spf-ip:log?
| | +--ro spf-ip:match?
| +--rw spf-ip:ipv4-pfes
| | +--rw spf-ip:ipv4-pfe [name]
| | +--rw spf-ip:name spf:spf-name-string
| | +--rw (remark-or-ipv4-pfe)?
| | +--:(remark)
| | | +--rw spf-ip:remark? spf:spf-remark
| | +--:(ipv4-pfe)
| | +--rw spf-ip:filters
| | | +--rw (source-address-host-group)
| | | | +--:(source-ip)
| | | | | +--rw spf-ip:ip-source-address
| | | | | +--rw spf-ip:ip-source-mask
| | | | +--:(ip-source-any)
| | | | | +--rw spf-ip:ip-source-any?
| | | | +--:(source-host)
| | | | | +--rw (ip-src-address-or-name)
| | | | | +--:(ip-source-host-address)
| | | | | | +--rw spf-ip:ip-source-host-address?
| | | | | +--:(ip-source-host-name)
| | | | | +--rw spf-ip:ip-source-host-name?
| | | | +--:(source-group)
| | | | +--rw spf-ip:ip-source-group?
| | | +--rw (dest-address-host-group)
| | | | +--:(dest-ip)
| | | | | +--rw spf-ip:ip-dest-address
| | | | | +--rw spf-ip:ip-dest-mask
| | | | +--:(ip-dest-any)
| | | | | +--rw spf-ip:ip-dest-any?
| | | | +--:(dest-host)
| | | | | +--rw (ip-dest-address-or-name)
| | | | | +--:(ip-dest-host-address)
| | | | | | +--rw spf-ip:ip-dest-host-address?
| | | | | +--:(ip-dest-host-name)
| | | | | +--rw spf-ip:ip-dest-host-name?
| | | | +--:(dest-group)
| | | | +--rw spf-ip:ip-dest-group?
| | | +--rw spf-ip:protocol?
| | | +--rw spf-ip:enable-capture?
| | | +--rw spf-ip:capture-session-id?
| | | +--rw spf-ip:fragments?
| | | +--rw spf-ip:time-range?
| | | +--rw (src-ports)?
| | | | +--:(port-number-range)
| | | | | +--rw spf-ip:src-port-lower
| | | | | +--rw spf-ip:src-port-upper
| | | | +--:(port-number)
Huang, et al. Expires March 08, 2014 [Page 28]
Internet-Draft yang-spf September 2013
| | | | | +--rw spf-ip:src-comparator
| | | | | +--rw spf-ip:src-port
| | | | +--:(port-group-ref)
| | | | +--rw spf-ip:src-port-group-name
| | | +--rw (dest-ports)?
| | | | +--:(port-number-range)
| | | | | +--rw spf-ip:des-port-lower
| | | | | +--rw spf-ip:des-port-upper
| | | | +--:(port-number)
| | | | | +--rw spf-ip:des-comparator
| | | | | +--rw spf-ip:des-port
| | | | +--:(port-group-ref)
| | | | +--rw spf-ip:des-port-group-name
| | | +--rw spf-ip:icmp-type?
| | | +--rw spf-ip:icmp-code?
| | | +--rw (packet-length-or-range)?
| | | | +--:(length)
| | | | | +--rw spf-ip:packet-length-comparator
| | | | | +--rw spf-ip:packet-length
| | | | +--:(range)
| | | | +--rw spf-ip:packet-length-upper
| | | | +--rw spf-ip:packet-length-lower
| | | +--rw spf-ip:tcp-flag-value?
| | | +--rw spf-ip:tcp-flag-mask?
| | | +--rw spf-ip:tcp-flag-operation?
| | | +--rw (ttl-value-or-range)?
| | | | +--:(value)
| | | | | +--rw spf-ip:ttl-comparator?
| | | | | +--rw spf-ip:ttl-value?
| | | | +--:(range)
| | | | +--rw spf-ip:ttl-value-lower?
| | | | +--rw spf-ip:ttl-value--upper?
| | | +--rw (dscp-or-tos)?
| | | +--:(dscp)
| | | | +--rw spf-ip:dscp?
| | | +--:(tos)
| | | +--rw spf-ip:tos?
| | | +--rw spf-ip:precedence?
| | +--rw spf-ip:actions
| | | +--rw spf-ip:action spf:spf-action
| | | +--rw spf-ip:log? empty
| | +--ro spf-ip:match? yang:counter64
| +--rw spf-ip:global-fragments? enumeration
| +--rw spf-mac:mac-pfes
| | +--rw spf-mac:mac-pfe [name]
| | +--rw spf-mac:name spf:spf-name-string
| | +--rw (remark-or-mac-pfe)?
| | +--:(remark)
Huang, et al. Expires March 08, 2014 [Page 29]
Internet-Draft yang-spf September 2013
| | | +--rw spf-mac:remark? spf:spf-remark
| | +--:(mac-pfe)
| | +--rw spf-mac:filters
| | | +--rw (source-network)
| | | | +--:(source-mac)
| | | | | +--rw spf-mac:source-address
| | | | | +--rw spf-mac:source-address-mask
| | | | +--:(source-any)
| | | | | +--rw spf-mac:source-any?
| | | | +--:(source-host)
| | | | +--rw (src-address-or-name)
| | | | +--:(source-host-address)
| | | | | +--rw spf-mac:source-host-address?
| | | | +--:(source-host-name)
| | | | +--rw spf-mac:source-host-name?
| | | +--rw (dest-network)
| | | | +--:(dest-mac)
| | | | | +--rw spf-mac:dest-address
| | | | | +--rw spf-mac:dest-address-mask
| | | | +--:(dest-any)
| | | | | +--rw spf-mac:dest-any?
| | | | +--:(dest-host)
| | | | +--rw (dest-address-or-name)
| | | | +--:(dest-host-address)
| | | | | +--rw spf-mac:dest-host-address?
| | | | +--:(dest-host-name)
| | | | +--rw spf-mac:dest-host-name?
| | | +--rw spf-mac:ethertype?
| | | +--rw spf-mac:ethertype-mask?
| | | +--rw spf-mac:cos?
| | | +--rw spf-mac:time-range?
| | | +--rw spf-mac:vlan?
| | | +--rw spf-mac:enable-capture?
| | | +--rw spf-mac:capture-session-id?
| | +--rw spf-mac:actions
| | | +--rw spf-mac:action
| | | +--rw spf-mac:log?
| | +--ro spf-mac:match?
| +--rw spf-arp:arp-pfes
| +--rw spf-arp:arp-pfe [name]
| +--rw spf-arp:name
| +--rw (remark-or-arp-pfe)?
| +--:(remark)
| | +--rw spf-arp:remark?
| +--:(arp-pfe)
| +--rw spf-arp:filters
| | +--rw spf-arp:direction?
| | +--rw (source-address-host-group)
Huang, et al. Expires March 08, 2014 [Page 30]
Internet-Draft yang-spf September 2013
| | | +--:(source-ip)
| | | | +--rw spf-arp:ip-source-address
| | | | +--rw spf-arp:ip-source-mask
| | | +--:(ip-source-any)
| | | | +--rw spf-arp:ip-source-any?
| | | +--:(source-host)
| | | | +--rw (ip-src-address-or-name)
| | | | +--:(ip-source-host-address)
| | | | | +--rw spf-arp:ip-source-host-address?
| | | | +--:(ip-source-host-name)
| | | | +--rw spf-arp:ip-source-host-name?
| | | +--:(source-group)
| | | +--rw spf-arp:ip-source-group?
| | +--rw (dest-address-host-group)
| | | +--:(dest-ip)
| | | | +--rw spf-arp:ip-dest-address
| | | | +--rw spf-arp:ip-dest-mask
| | | +--:(ip-dest-any)
| | | | +--rw spf-arp:ip-dest-any?
| | | +--:(dest-host)
| | | | +--rw (ip-dest-address-or-name)
| | | | +--:(ip-dest-host-address)
| | | | | +--rw spf-arp:ip-dest-host-address?
| | | | +--:(ip-dest-host-name)
| | | | +--rw spf-arp:ip-dest-host-name?
| | | +--:(dest-group)
| | | +--rw spf-arp:ip-dest-group?
| | +--rw (source-network)
| | | +--:(source-mac)
| | | | +--rw spf-arp:source-address
| | | | +--rw spf-arp:source-address-mask
| | | +--:(source-any)
| | | | +--rw spf-arp:source-any?
| | | +--:(source-host)
| | | +--rw (src-address-or-name)
| | | +--:(source-host-address)
| | | | +--rw spf-arp:source-host-address?
| | | +--:(source-host-name)
| | | +--rw spf-arp:source-host-name?
| | +--rw (dest-network)
| | | +--:(dest-mac)
| | | | +--rw spf-arp:dest-address
| | | | +--rw spf-arp:dest-address-mask
| | | +--:(dest-any)
| | | | +--rw spf-arp:dest-any?
| | | +--:(dest-host)
| | | +--rw (dest-address-or-name)
| | | +--:(dest-host-address)
Huang, et al. Expires March 08, 2014 [Page 31]
Internet-Draft yang-spf September 2013
| | | | +--rw spf-arp:dest-host-address?
| | | +--:(dest-host-name)
| | | +--rw spf-arp:dest-host-name?
| | +--rw spf-arp:enable-capture?
| | +--rw spf-arp:capture-session-id?
| +--rw spf-arp:actions
| | +--rw spf-arp:action
| | +--rw spf-arp:log?
| +--ro spf-arp:match?
+--rw port-groups
| +--rw port-group [name]
| +--rw name
| +--rw port-group-entry [name]
| +--rw name
| +--rw (port-number-or-range)?
| +--:(port-number-range)
| | +--rw port-lower
| | +--rw port-upper
| +--:(port-number)
| +--rw comparator
| +--rw port
+--rw timerange-groups
| +--rw timerange-group [name]
| +--rw name
| +--rw time-range [name]
| +--rw name
| +--rw remark?
| +--rw (range-type)?
| +--:(absolute)
| | +--rw absolute
| | +--rw start?
| | +--rw end?
| +--:(periodic)
| +--rw periodic
| +--rw weekdays?
| +--rw start?
| +--rw end?
+--rw ip-address-groups
+--rw ip-address-group [name]
+--rw name
+--rw afi?
+--rw ip-address [name]
+--rw name
+--rw (ip-network-kind)
+--:(ip)
| +--rw ip-address?
| +--rw ip-mask
+--:(ip-any)
Huang, et al. Expires March 08, 2014 [Page 32]
Internet-Draft yang-spf September 2013
| +--rw ip-any?
+--:(host)
+--rw (address-or-name)
+--:(ip-host-address)
| +--rw ip-host-address?
+--:(ip-host-name)
+--rw ip-host-name?
module: spf-ip
module: spf-mac
module: spf-arp
9. SPF Examples
9.1. Configuration Example
Requirement: Denies TELNET traffic from 14.3.6.234 bound for host
6.5.4.1 from leaving. Denies all TFTP traffic bound for TFTP
servers. Permits all other IP traffic.
In order to achieve the requirement, an name access control list is
needed. In the spf, we need three pfes. The spf and pfes can be
described in CLI: as the following:
access-list ip ispf
deny tcp 14.3.6.234 0.0.0.0 host 6.5.4.1 eq 23
deny udp any any eq tftp
permit ip any any
Here is the example spf configuration xml:
<rpc message-id="101"
xmlns:nc="urn:cisco:params:xml:ns:yang:spf:1.0"
xmlns:spf-ip="urn:cisco:params:xml:ns:yang:spf-ip"
// replace with IANA namespace when assigned
<edit-config>
<target>
<running/>
</target>
<config>
<top xmlns="http://example.com/schema/1.2/config">
<spfs>
<spf >
<name>sample-ip-spf</name>
Huang, et al. Expires March 08, 2014 [Page 33]
Internet-Draft yang-spf September 2013
<spf-type>ip-spf</spf-type>
<enable-match-counter>false</enable-match-counter>
<spf-ip:afi>ipv4</spf-ip:afi>
<spf-ip:ipv4-pfes>
<spf-ip:ipv4-pfe>
<spf-ip:name>pfe10</spf-ip:name>
<spf-ip:filters>
<spf-ip:protocol>6</spf-ip:protocol>
<spf-ip:ip-source-address>
14.3.6.234
</spf-ip:ip-source-address>
<spf-ip:ip-source-mask>0.0.0.0</spf-ip:ip-source-mask>
<spf-ip:ip-dest-host-address>
6.5.4.1
</spf-ip:ip-dest-host-address>
<spf-ip:des-comparator>eq</spf-ip:des-comparator>
<spf-ip:des-port>23</spf-ip:des-port>
</spf-ip:filters>
<spf-ip:actions>
<spf-ip:action>deny</spf-ip:action>
</spf-ip:actions>
</spf-ip:ipv4-pfe>
<spf-ip:ipv4-pfe>
<spf-ip:name>pfe20</spf-ip:name>
<spf-ip:filters>
<spf-ip:protocol>17</spf-ip:protocol>
<spf-ip:ip-source-any/>
<spf-ip:ip-dest-any/>
<spf-ip:des-comparator>eq</spf-ip:des-comparator>
<spf-ip:des-port>69</spf-ip:des-port>
</spf-ip:filters>
<spf-ip:actions>
<spf-ip:action>deny</spf-ip:action>
</spf-ip:actions>
</spf-ip:ipv4-pfe>
<spf-ip:ipv4-pfe>
<spf-ip:name>pfe30</spf-ip:name>
<spf-ip:filters>
<spf-ip:ip-source-any/>
<spf-ip:ip-dest-any/>
</spf-ip:filters>
<spf-ip:actions>
<spf-ip:action>permit</spf-ip:action>
</spf-ip:actions>
</spf-ip:ipv4-pfe>
Huang, et al. Expires March 08, 2014 [Page 34]
Internet-Draft yang-spf September 2013
</spf-ip:ipv4-pfes>
</spf>
</spfs>
</top>
</config>
</edit-config>
</rpc>
10. Stateless-PF YANG Module
This module imports type definitions from [RFC6021].
<CODE BEGINS> file "stateless-pf@2013-09-03.yang"
module stateless-pf {
namespace "urn:cisco:params:xml:ns:yang:spf";
// replace with IANA namespace when assigned
prefix spf;
import ietf-inet-types {
prefix "inet";
}
import ietf-yang-types {
prefix "yang";
}
organization
"IETF NETMOD (NETCONF Data Modeling Language) Working Group";
contact
"WG Web: http://tools.ietf.org/wg/netmod/
WG List: netmod@ietf.org
WG Chair: David Kessens
david.kessens@nsn.com
WG Chair: Juergen Schoenwaelder
j.schoenwaelder@jacobs-university.de
Editor: Lisa Huang
yihuan@cisco.com
Editor: Alexander Clemm
alex@cisco.com
Huang, et al. Expires March 08, 2014 [Page 35]
Internet-Draft yang-spf September 2013
Editor: Andy Bierman
andy@yumaworks.com";
description
"This YANG module defines a component that describing the
configuration of Stateless Packet Filters (SPF), also known as
Access Control Lists (SPFs).
An SPF is an ordered set of rules and actions used to filter
traffic. Each set of rules and actions is represented
as an Packet Filter Entry (PFE), also known as Access
Control Entries (PFE). Each PFE is evaluated
sequentially. When the rule matches then action for that
rule is applied to the packet.
There are three types of SPF.
IP SPFs - IP SPFs are ordered sets of rules that can use to
filter traffic based on IP information in the Layer 3
header of packets.
The device applies IP SPFs only to IP traffic. IP SPF
can be IPv4 or IPv6.
MAC SPFs - MAC SPFs are used to filter traffic using the
information in the Layer 2 header of each packet.
MAC SPFs are by default only applied to non-IP
traffic; however, Layer 2 interfaces can be configured
to apply MAC SPFs to all traffic.
ARP SPFs - The device applies ARP SPFs to IP traffic.
This module should be used with spf-ip, spf-arp, or spf-mac
depends on what feature the device supports.
This YANG module also includes auxiliary definitions that
are needed in conjunction with configuration of SPFs, such as
reusable containers and references for ports and IP.
Terms and Acronyms
PFE (pfe): Packet Filter Entry
SPF (spf): Stateless Packet Filter
AFI (afi): Authority and Format Identifier (Address
Field Identifier)
ARP (arp): Address Resolution Protocol
IP (ip): Internet Protocol
Huang, et al. Expires March 08, 2014 [Page 36]
Internet-Draft yang-spf September 2013
IPv4 (ipv4):Internet Protocol Version 4
IPv6 (ipv6): Internet Protocol Version 6
MAC: Media Access Control
TCP (tcp): Transmission Control Protocol
TTL (ttl): Time to Live
VLAN (vlan): Virtual Local Area Network
";
revision 2013-09-03 {
description "Initial revision. ";
}
/* Features */
feature capture-session-id {
if-feature packet-capture;
description
"The ability to configure SPF capture in order to
selectively monitor traffic on an interface or VLAN.
When the capture option for an SPF rule
is enabled, packets that match this rule are
either forwarded or dropped based on the specified permit
or deny action and may also be copied to an alternate
destination port for further analysis.
An SPF rule with the capture option can be applied
as follows:
On a VLAN
In the ingress direction on all interfaces
In the egress direction on all Layer 3 interfaces
The statistics data for the capture-session are capture
in the device where the SPF rule applied to.";
}
feature host-by-name {
description
"The capability to reference a host by DNS name.";
}
feature ip-address-groups {
description
"The ability to define named groups for lists of
ip addresses. ";
}
Huang, et al. Expires March 08, 2014 [Page 37]
Internet-Draft yang-spf September 2013
feature logging {
description
"The ability to log messages upon the matching of SPFs.";
}
feature match-counter {
description
"The ability to maintain global or local match statistics
for each SPF rules.";
}
feature packet-capture {
description "The ability to capture packets that
match the filter.";
}
feature packet-length {
description "The ability to filter packets by packet length";
}
feature port-groups {
description
"The ability to define named groups for lists of ports. ";
}
/* Identities */
identity spf-type {
description "Base spf type for all SPF type identifiers.";
}
/* Types */
typedef spf-comparator {
description "A data type used to express comparator string";
type enumeration {
enum "eq" {
value 0;
description "match only equal to any giving number.";
}
enum "gt" {
value 1;
description
"match only greater than any giving number.";
}
enum "lt" {
Huang, et al. Expires March 08, 2014 [Page 38]
Internet-Draft yang-spf September 2013
value 2;
description
"match only lower than any giving number.";
}
enum "neq" {
value 3;
description
"match only not equal to any giving number";
}
}
}
typedef spf-action {
description "An enumeration data type to express spf
action when match.";
type enumeration {
enum deny {
description "Apply deny action to the traffic";
}
enum permit {
description "Apply permit action to the traffic";
}
}
}
typedef spf-remark {
type string {
length "0..100";
}
description
"A remark is a comment that can be
associated with an PFE in order to make
the access list easier for the network
administrator to understand.
It is retained to facilitate
co-existence with CLI.";
}
typedef spf-type-ref {
description
"This type is used to refer to an Stateless Packet Filter
(spf) type";
type identityref {
base "spf-type";
}
}
Huang, et al. Expires March 08, 2014 [Page 39]
Internet-Draft yang-spf September 2013
typedef spf-ref {
description "This type refers to an SPF.";
type leafref {
path "/spf:spfs/spf:spf/spf:name";
}
}
typedef port-group-ref {
description
"This type is used to refer to a Portgroup object.";
type leafref {
path "/spfs/port-groups/port-group/name";
}
}
typedef ip-address-group-ref {
description
"This type is used to refer to a time range object.";
type leafref {
path "/spfs/ip-address-groups/ip-address-group/name";
}
}
typedef time-range-ref {
description
"This type is used to refer to a time range object.";
type leafref {
path "/spfs/timerange-groups/timerange-group/name";
}
}
typedef weekdays {
type bits {
bit Sunday {
position 0;
}
bit Monday {
position 1;
}
bit Tuesday {
position 2;
}
bit Wednesday {
position 3;
}
bit Thursday {
Huang, et al. Expires March 08, 2014 [Page 40]
Internet-Draft yang-spf September 2013
position 4;
}
bit Friday {
position 5;
}
bit Saturday {
position 6;
}
}
}
typedef spf-name-string {
type string {
length "1 .. 64";
}
}
/* Groupings */
grouping PFE-COMMON {
description
"A collection of nodes that should be added to
every PFE list entry";
container actions {
leaf action {
type spf:spf-action;
mandatory true;
description "Permit/deny action.";
}
leaf log {
if-feature spf:logging;
type empty;
description
"Causes an informational logging message about the
packet that matches the entry to be sent to the
console.";
}
}
leaf match {
if-feature spf:match-counter;
config false;
type yang:counter64;
description
"The total packet that have matched for the
particular PFE";
Huang, et al. Expires March 08, 2014 [Page 41]
Internet-Draft yang-spf September 2013
}
}
grouping FILTER-COMMON {
description
"A collection of nodes that should be added to
every 'filters' container within each
PFE list entry";
leaf enable-capture {
if-feature spf:packet-capture;
type boolean;
description
"Enable packet capture on this filter
for this session.";
}
leaf capture-session-id {
if-feature spf:capture-session-id;
when "../enable-capture = 'true'";
type uint32 {
range "1..48";
}
description
"Enable packet capture on this filter
for this session id.";
}
}
/* Data Nodes */
container spfs {
description
"This is the top container that contains a list of
named SPF and reusable spf object groups.";
list spf {
key name;
leaf name {
description "spf/access group name.";
type spf-name-string;
}
leaf spf-type {
type spf-type-ref;
description "Type of SPF";
mandatory true;
}
leaf enable-capture-global {
Huang, et al. Expires March 08, 2014 [Page 42]
Internet-Draft yang-spf September 2013
if-feature packet-capture;
type boolean;
description "Enable packet capture on this filter
for this session. Session ID range is 1 to 48";
default "false";
}
leaf capture-session-id-global {
if-feature capture-session-id;
when "../enable-capture-global = 'true'";
type uint32 {
range "1..48";
}
description "Enable packet capture on this filter
for this session. Session ID range is 1 to 48";
}
choice enable-match-counter-choices {
if-feature match-counter;
case match {
leaf enable-match-counter {
type boolean;
description
"Enable to collect statistics for the SPF";
default false;
}
}
case per-entry-match {
leaf enable-per-entry-match-counter {
type boolean;
description "Enable to collect match
statistics for each SPF entry(Stateless PFE).";
default false;
}
}
}
leaf match {
if-feature match-counter;
config false;
type yang:counter64;
description
"The total packet that have matched for the
particular access list";
}
}
container port-groups {
if-feature port-groups;
Huang, et al. Expires March 08, 2014 [Page 43]
Internet-Draft yang-spf September 2013
list port-group {
key "name";
leaf name {
type spf-name-string;
}
list port-group-entry {
key "name";
ordered-by user;
leaf name {
type spf-name-string;
}
//unique "comparator port-number
//port-lower port-upper";
choice port-number-or-range {
case port-number-range {
description
"Port group includes all ports between
port-lowerand port-upper (including those)";
leaf port-lower {
type inet:port-number;
description "Lower Port number.";
mandatory true;
}
leaf port-upper {
type inet:port-number;
description "Upper Port number.";
mandatory true;
must "../port-lower <= ../port-upper";
}
}
case port-number {
description
"Port group includes all ports that are greater
than, greater or equal, less than, less or
equal, or not equal the port, per the
indicated comparator.
It is possible for the port group to be empty
(for example, in case a port group that
is less than the minimum port number is
specified).";
leaf comparator {
type spf-comparator;
mandatory true;
}
leaf port {
type inet:port-number;
description "Port number.";
Huang, et al. Expires March 08, 2014 [Page 44]
Internet-Draft yang-spf September 2013
mandatory true;
}
}
} // choice port-number-or-range
} // list port-group-entry
} // list port-group
} // container port-groups
container timerange-groups {
description "Define time range entries to restrict
the access. The time range is identified by a name
and then referenced by a function, so that those
time restrictions are imposed on the function itself.";
list timerange-group {
key "name";
leaf name {
type spf-name-string;
}
list time-range {
key "name";
ordered-by user;
leaf name {
type spf-name-string;
}
leaf remark {
type spf-remark;
}
choice range-type {
// abosolute or periodic time range
container absolute {
description
"Absolute time and date that
the associated function starts
going into effect.";
leaf start {
type yang:date-and-time;
description
"Absolute start time and date";
}
leaf end {
type yang:date-and-time;
description "Absolute end time and date";
}
}
container periodic {
description
Huang, et al. Expires March 08, 2014 [Page 45]
Internet-Draft yang-spf September 2013
"To specify a periodic time and date.";
leaf weekdays {
type weekdays;
}
leaf start {
type yang:timestamp;
description "Start time";
}
leaf end {
type yang:timestamp;
description "End time";
}
}
} // choice range-type
} // list time-range
} // list timerange-group
} // container timerange-groups
container ip-address-groups {
if-feature ip-address-groups;
description
"This contains a list of named ip address group. Each
group defines a range of address and mask pair.";
list ip-address-group {
key "name";
leaf name {
type spf-name-string;
}
leaf afi {
default "ipv4";
type inet:ip-version;
description "Address Field Identifier (AFI).";
}
list ip-address {
key "name";
ordered-by user;
leaf name {
type spf-name-string;
}
//unique "ip-address ip-mask";
//unique "ip-host-address";
grouping IP-HOST {
description
"Choice within a case not allowed so need
this grouping.";
choice address-or-name {
mandatory true;
Huang, et al. Expires March 08, 2014 [Page 46]
Internet-Draft yang-spf September 2013
leaf ip-host-address {
type inet:ip-address;
}
leaf ip-host-name {
if-feature spf:host-by-name;
type inet:domain-name;
}
}
}
choice ip-network-kind {
mandatory true;
case ip {
leaf ip-address {
type inet:ip-address;
}
leaf ip-mask {
type inet:ip-prefix;
mandatory true;
}
}
leaf ip-any {
type empty;
description "To express Any network or address.
Use the any keyword as an abbreviation
for an address and a mask of 0.0.0.0
255.255.255.255. For example:
0.0.0.0/255.255.255.255 means 'any'";
}
case host {
description
"Use the host address combination as an
abbreviation for an address and wildcard
of address 0.0.0.0";
uses IP-HOST;
}
// case group not allowed here!
}
} // list ip-address
} // list ip-address-group
} // container ip-address-groups
} // container spfs
}
Huang, et al. Expires March 08, 2014 [Page 47]
Internet-Draft yang-spf September 2013
<CODE ENDS>
11. SPF-IP YANG Module
This module imports type definitions from [RFC6021] and common-types
yang defined with stateless-pf model.
<CODE BEGINS> file "spf-ip@2013-09-03.yang"
module spf-ip {
namespace "urn:cisco:params:xml:ns:yang:spf-ip";
// replace with IANA namespace when assigned
prefix spf-ip;
import stateless-pf {
prefix spf;
}
import ietf-inet-types {
prefix "inet";
}
import common-types {
prefix "c-types";
}
organization
"IETF NETMOD (NETCONF Data Modeling Language) Working Group";
contact
"WG Web: http://tools.ietf.org/wg/netmod/
WG List: netmod@ietf.org
WG Chair: David Kessens
david.kessens@nsn.com
WG Chair: Juergen Schoenwaelder
j.schoenwaelder@jacobs-university.de
Editor: Lisa Huang
yihuan@cisco.com
Editor: Alexander Clemm
alex@cisco.com
Editor: Andy Bierman
andy@yumaworks.com";
description
"This YANG module augments the 'stateless-pf' module with configuration
Huang, et al. Expires March 08, 2014 [Page 48]
Internet-Draft yang-spf September 2013
and operational data for IPv4 and IPv6 stateless
packet filter.
An Stateless Packet Filter (SPF), also know as an Access
Control List (SPF), is an ordered set of rules and
actions used to filter traffic.
Each set of rules and actions is represented as a Packet Filter
Entry (PFE), also know as an Access
Control Entries (PFE). Each PFE is evaluated sequentially.
When the rule matches then action for that rule is applied
to the packet.
IP SPFs are ordered sets of rules that can use to
filter traffic based on IP information in the Layer 3 header
of packets.
The device applies IP SPFs only to IP traffic. IP SPF
can be IPv4 or IPv6.
Terms and Acronyms
PFE (pfe): Packet Filter Entry
SPF (spf): Stateless Packet Filter
AFI (afi): Authority and Format Identifier (Address Field
Identifier)
DSCP (dscp): Differentiated Services Code Point
ICMP (icmp): Internet Control Message Protocol
IGMP (igmp): Internet Group Management Protocol
IP (ip): Internet Protocol
IPv4 (ipv4):Internet Protocol Version 4
IPv6 (ipv6): Internet Protocol Version 6
QoS: Quality of Service
TCP (tcp): Transmission Control Protocol
ToS (tos): Type of Service
TTL (ttl): Time to Live
UDP (udp): User Datagram Protocol
Huang, et al. Expires March 08, 2014 [Page 49]
Internet-Draft yang-spf September 2013
VLAN (vlan): Virtual Local Area Network
VRF(vrf) : Virtual Routing and Forwarding
";
revision 2013-09-03 {
description "Initial revision. ";
}
/* Features */
feature time-to-live {
description "The ability to filter packets based on their
time-to-live (TTL) value (0 to 255)";
reference "SPF Support for Filtering on TTL Value";
}
feature flow-label {
description
"The ability to filter packets based on flow lable.
The 20-bit Flow Label field in the IPv6 header
is used by a source to label packets
of a flow. This is an IPv6 PFEs option.";
reference "RFC 3697 IPv6 Flow Label Specification";
}
/* Identities */
identity ip-spf {
base "spf:spf-type";
description "layer 3 SPF type";
}
/* Groupings */
grouping IP-SOURCE-NETWORK {
description "Reusable IP address and mask pair.";
grouping IP-SOURCE-HOST {
description
"Choice within a case not allowed so need
this grouping.";
choice ip-src-address-or-name {
mandatory true;
leaf ip-source-host-address {
type inet:ip-address;
}
Huang, et al. Expires March 08, 2014 [Page 50]
Internet-Draft yang-spf September 2013
leaf ip-source-host-name {
if-feature spf:host-by-name;
type inet:domain-name;
}
}
}
choice source-address-host-group {
mandatory true;
case source-ip {
description "Used with address and mask couple
to express network.";
leaf ip-source-address {
type inet:ip-address;
mandatory true;
}
leaf ip-source-mask {
type inet:ip-address;
mandatory true;
}
}
leaf ip-source-any {
type empty;
description "To express Any network or address.
Use the any keyword as an abbreviation
for an address and a mask of 0.0.0.0
255.255.255.255. For example:
0.0.0.0/255.255.255.255 means 'any'";
}
case source-host {
description "Used with host address to express a
single host
Use the host address(or name)
combination is the same as an address
and mask of address 0.0.0.0.
For example: '10.1.1.2/0.0.0.0' is the same
as 'host 10.1.1.2'";
uses IP-SOURCE-HOST;
}
case source-group {
if-feature spf:ip-address-groups;
leaf ip-source-group {
type spf:ip-address-group-ref;
}
}
}
}
Huang, et al. Expires March 08, 2014 [Page 51]
Internet-Draft yang-spf September 2013
grouping IP-DESTINATION-NETWORK {
description
"Reusable IP address and mask pair for destination.";
grouping IP-DESTINATION-HOST {
description
"Choice within a case not allowed so need
this grouping.";
choice ip-dest-address-or-name {
mandatory true;
leaf ip-dest-host-address {
type inet:ip-address;
}
leaf ip-dest-host-name {
if-feature spf:host-by-name;
type inet:domain-name;
}
}
}
choice dest-address-host-group {
mandatory true;
case dest-ip {
description "Used with address and mask couple
to express network.";
leaf ip-dest-address {
type inet:ip-address;
mandatory true;
}
leaf ip-dest-mask {
type inet:ip-address;
mandatory true;
}
}
leaf ip-dest-any {
type empty;
description "To express Any network or address.
Use the any keyword as an abbreviation
for an address and a mask of 0.0.0.0
255.255.255.255. For example:
0.0.0.0/255.255.255.255 means 'any'";
}
case dest-host {
description "Used with host address to express a
single host
Use the host address(or name)
combination is the same as an address
and mask of address 0.0.0.0.
Huang, et al. Expires March 08, 2014 [Page 52]
Internet-Draft yang-spf September 2013
For example: '10.1.1.2/0.0.0.0' is the same
as 'host 10.1.1.2'";
uses IP-DESTINATION-HOST;
}
case dest-group {
if-feature spf:ip-address-groups;
description "Use the group keyword and group name
to refer to a pre-defined address object group
which is a list of address and mask.";
leaf ip-dest-group {
type spf:ip-address-group-ref;
}
}
}
}
grouping DSCP-OR-TOS {
choice dscp-or-tos {
leaf dscp {
type inet:dscp;
description
"Match packets with given dscp value";
}
case tos {
leaf tos {
type c-types:tos;
description
"Match packets with given TOS value";
}
leaf precedence {
when "boolean(../tos)" ;
type c-types:precedence;
description
"Match packets with given precedence value";
}
}
}
}
grouping IP-PFE-FILTERS {
leaf protocol {
type c-types:ip-protocol;
description "IP protocol number.";
}
Huang, et al. Expires March 08, 2014 [Page 53]
Internet-Draft yang-spf September 2013
uses spf:FILTER-COMMON;
leaf fragments {
type empty;
description "Check non-initial fragments";
}
leaf time-range {
type spf:time-range-ref;
description
"Refer a time range object by
name (Max Size 64).";
}
choice src-ports {
when "protocol = '6' or protocol = '17' or " +
"protocol = '132'";
description
"Apply only when the protocol is TCP,
UDP or SCTP.";
case port-number-range {
description
"Port group includes all ports between port-lower
and port-upper (including those)";
leaf src-port-lower {
type inet:port-number;
description "Lower Port number.";
mandatory true;
}
leaf src-port-upper {
type inet:port-number;
description "Upper Port number.";
mandatory true;
must "../src-port-lower <= ../src-port-upper";
}
}
case port-number {
description
"Port group includes all ports that are greater
than, greater or equal, less than, less or equal,
or not equal the port, per the indicated
comparator. It is possible for the port group
to be empty (for example, in case a port group
that is less than the minimum port number is
specified).";
leaf src-comparator {
Huang, et al. Expires March 08, 2014 [Page 54]
Internet-Draft yang-spf September 2013
type spf:spf-comparator;
mandatory true;
}
leaf src-port {
type inet:port-number;
description "Port number.";
mandatory true;
}
}
case port-group-ref {
if-feature spf:port-groups;
leaf src-port-group-name {
type spf:port-group-ref;
mandatory true;
description
"Reference a port group by the Port
Group name.";
}
}
} // choice src-ports
choice dest-ports {
when "protocol = '6' or protocol = '17' or " +
"protocol = '132'";
description
"Apply only when the protocol is TCP,
UDP or SCTP.";
case port-number-range {
description "Port group includes all ports between
port-lower and port-upper (including those)";
leaf des-port-lower {
type inet:port-number;
description "Lower Port number.";
mandatory true;
}
leaf des-port-upper {
type inet:port-number;
description "Upper Port number.";
mandatory true;
must "../des-port-lower <= ../des-port-upper";
}
}
case port-number {
description "Port group includes all ports that
are greater than, greater or equal, less than,
less or equal, or not equal the port, per the
indicated comparator. It is possible for the
Huang, et al. Expires March 08, 2014 [Page 55]
Internet-Draft yang-spf September 2013
port group to be empty (for example, in case a
port group that is less than the minimum port
number is specified).";
leaf des-comparator {
type spf:spf-comparator;
mandatory true;
}
leaf des-port {
type inet:port-number;
description "Port number.";
mandatory true;
}
}
case port-group-ref {
if-feature spf:port-groups;
leaf des-port-group-name {
type spf:port-group-ref;
mandatory true;
description
"Reference a port group by the Port Group name.";
}
}
} // choice dest-ports
leaf icmp-type {
when "../protocol = '1'";
type c-types:icmp-type;
description
"ICMP message type number.
Apply only when the protocol is icmp";
}
leaf icmp-code {
when "boolean(../icmp-type) ";
type c-types:icmp-code;
description
"ICMP subtype for a given icmp type.";
}
choice packet-length-or-range {
if-feature spf:packet-length;
case length {
leaf packet-length-comparator {
type spf:spf-comparator;
description
"Operant that compare the packet
length. Operands are lt (less than),
gt (greater than), eq (equal), and neq
Huang, et al. Expires March 08, 2014 [Page 56]
Internet-Draft yang-spf September 2013
(not equal).";
mandatory true;
}
leaf packet-length {
type uint32 {
range "20..9210";
}
description
"Packet length value for
operation gt, eq, etc, other
than range";
//TODO need to find out why package is
// less than 9210
mandatory true;
}
}
case range {
description
"Packet operator 'range' takes
both lower and upper value.";
leaf packet-length-upper {
type uint32 {
range "20..9210";
}
mandatory true;
description "Upper Packet length";
}
leaf packet-length-lower {
type uint32 {
range "20..9210";
}
must "number(../packet-length-lower) <= " +
"number(../packet-length-upper)";
mandatory true;
description "Lower packet length";
}
}
}
leaf tcp-flag-value {
type c-types:tcp-flag-type ;
description "TCP flag bits that needs to be checked";
}
leaf tcp-flag-mask {
when "boolean(../tcp-flag-value)" ;
Huang, et al. Expires March 08, 2014 [Page 57]
Internet-Draft yang-spf September 2013
type c-types:tcp-flag-type ;
description "TCP flag bit that needs to be checked";
}
leaf tcp-flag-operation {
when "boolean(../tcp-flag-value)" ;
description
"TCP flag Match option.
A match occurs if the TCP
datagram has certain TCP flags
set or not set. You use the
match-any keyword to allow a match
to occur if any of the specified
TCP flags are present, or you can
use the match-all keyword to allow
a match to occur only if all of
the specified TCP flags are
present. You must follow the
match-any and match-all keywords
with the + or - keyword and the
flag-name argument to match on
one or more TCP flags. ";
default match-any;
type enumeration {
enum match-any {
description "match any";
}
enum match-all {
description "match all";
}
}
}
choice ttl-value-or-range {
if-feature time-to-live;
case value {
leaf ttl-comparator {
type spf:spf-comparator;
description
"Compares the TTL value in the packet
to the TTL value specified in this
PFE statement. Operands are lt (less
than), gt (greater than), and eq
(equal), neq (not equal).";
}
leaf ttl-value {
type c-types:time-to-live;
Huang, et al. Expires March 08, 2014 [Page 58]
Internet-Draft yang-spf September 2013
}
}
case range {
leaf ttl-value-lower {
type c-types:time-to-live;
description "Lower ttl number.";
}
leaf ttl-value--upper {
type c-types:time-to-live;
description "Upper ttl number.";
}
}
}
}
/* Data Nodes */
augment "/spf:spfs/spf:spf" {
when "spf:spf-type = 'ip-spf'";
leaf afi {
type inet:ip-version ;
default "ipv4";
}
container ipv6-pfes {
when "../afi = 'ipv6'" ;
description
" The ip-pfes container contains a list of ip-pfe.
Each ip-pfe is made of a unique ID, an optional
remark (comment), and a filter. The filter
requires a mandatory action (permit/deny) and one or
more options such as source-address with mask,ttl etc";
list ipv6-pfe {
key "name";
ordered-by user;
description "Layer 3 Packet Filter Entry (PFE)";
leaf name {
type spf:spf-name-string;
description "Unique PFE identifier.";
}
choice remark-or-ipv6-case {
leaf remark {
Huang, et al. Expires March 08, 2014 [Page 59]
Internet-Draft yang-spf September 2013
type spf:spf-remark;
// mandatory true;
}
case ipv6-pfe {
container filters {
uses IP-SOURCE-NETWORK;
uses IP-DESTINATION-NETWORK;
uses IP-PFE-FILTERS;
uses DSCP-OR-TOS;
leaf igmp-type {
when "../protocol = '2' ";
type c-types:igmp-code;
description
"IGMP message type (0 to 15) for
filtering IGMP packets. Apply only
when the protocol is igmp in ipv4";
}
leaf flow-label {
if-feature flow-label;
when "../protocol = '17'";
type uint64 {
range "0..1048575";
}
description
"Flow label value. Apply only when
the protocol is UDP in ipv6.";
reference
"RFC3697 IPv6 Flow Label Specification";
}
} // container filters
uses spf:PFE-COMMON;
} // case ipv6-pfe
} // choice remark-or-ipv6-pfe
} // list ipv6-pfe
} // container ipv6-pfes
container ipv4-pfes {
when "../afi = 'ipv4'" ;
description
"The ip-pfes container contains a list of ip-pfe.
Each ip-pfe is made of a unique ID, an optional
remark (comment), and a filter. The filter requires a
mandatory action (permit/deny) and one or more options
Huang, et al. Expires March 08, 2014 [Page 60]
Internet-Draft yang-spf September 2013
such as source-address with mask,ttl etc";
list ipv4-pfe {
key "name";
ordered-by user;
description "Layer 3 Packet Filter Entry (PFE)";
leaf name {
type spf:spf-name-string;
description "Unique PFE identifier";
}
choice remark-or-ipv4-pfe {
leaf remark {
type spf:spf-remark;
// mandatory true;
}
case ipv4-pfe {
container filters {
uses IP-SOURCE-NETWORK;
uses IP-DESTINATION-NETWORK;
uses IP-PFE-FILTERS;
uses DSCP-OR-TOS;
}
uses spf:PFE-COMMON;
} // case ipv4-pfe
} // choice remark-or-ipv4-pfe
} // list ipv4-pfe
} // container ipv4-pfes
leaf global-fragments {
default "not-set";
type enumeration {
enum not-set;
enum permit-all {
description "Allow all fragments";
}
enum deny-all {
description "Drop all fragments";
}
}
description
"Optimizes fragment handling for noninitial fragments.
When this leaf is set to 'permit-all', noninitial
fragments will be permitted unless explicitly denied.
When this leaf is set to 'deny-all', noninitial
fragments will be denied unless explicitly
permitted. ";
Huang, et al. Expires March 08, 2014 [Page 61]
Internet-Draft yang-spf September 2013
}
}
}
<CODE ENDS>
12. SPF-MAC Configuration YANG Module
This module imports type definitions from common-types YANG defined
in this model.
<CODE BEGINS> file "spf-mac@2013-09-03.yang"
module spf-mac {
namespace "urn:cisco:params:xml:ns:yang:spf-mac";
// replace with IANA namespace when assigned
prefix spf-mac;
import stateless-pf { prefix spf; }
import common-types {
prefix "c-types";
}
import ietf-inet-types {
prefix "inet";
}
import ietf-yang-types {
prefix "yang";
}
organization
"IETF NETMOD (NETCONF Data Modeling Language) Working Group";
contact
"WG Web: http://tools.ietf.org/wg/netmod/
WG List: netmod@ietf.org
WG Chair: David Kessens
david.kessens@nsn.com
WG Chair: Juergen Schoenwaelder
j.schoenwaelder@jacobs-university.de
Huang, et al. Expires March 08, 2014 [Page 62]
Internet-Draft yang-spf September 2013
Editor: Lisa Huang
yihuan@cisco.com
Editor: Alexander Clemm
alex@cisco.com
Editor: Andy Bierman
andy@yumaworks.com";
description
"This YANG module augments the 'stateless-pf' module with
configuration and operational data for MAC stateless packet
filter.
An Stateless Packet Filter (SPF), also know as an Access
Control List (SPF), is an ordered set of rules and
actions used to filter traffic.
Each set of rules and actions is represented as a Packet Filter
Entry (PFE), also know as an Access
Control Entries (PFE). Each PFE is evaluated sequentially.
When the rule matches then action for that rule is applied
to the packet.
MAC SPFs - MAC SPFs are used to filter traffic using the
information in the Layer 2 header of each packet.
MAC SPFs are by default only applied to non-IP
traffic; however, Layer 2 interfaces can be configured to
apply MAC SPFs to all traffic.
Terms and Acronyms
PFE (pfe): Packet FIlter Entry
SPF (spf): Stateless Packet Filter
AFI (afi): Authority and Format Identifier (Address Field
Identifier)
CoS (cos): Class of Service
MAC: Media Access Control
TTL (ttl): Time to Live
VLAN (vlan): Virtual Local Area Network
VRF(vrf) : Virtual Routing and Forwarding
";
Huang, et al. Expires March 08, 2014 [Page 63]
Internet-Draft yang-spf September 2013
revision 2013-09-03 {
description "Initial revision. ";
}
/* Features */
feature ethertype-mask {
description
"The ability to fiter packets based on ether-type mask
in hex 0x0-0xFFFF.";
}
/* Identities */
identity mac-spf {
base spf:spf-type;
description "layer 2 SPF type";
}
/* Groupings */
grouping MAC-SOURCE-NETWORK {
description "MAC address and mask pair for source.";
grouping MAC-SOURCE-HOST {
description
"Choice within a case not allowed so need
this grouping.";
choice src-address-or-name {
mandatory true;
leaf source-host-address {
type inet:ip-address;
description
"Use the host address combination as an
abbreviation for an address and wildcard
of address 0.0.0.0";
}
leaf source-host-name {
if-feature spf:host-by-name;
type inet:domain-name;
}
}
}
choice source-network {
mandatory true;
case source-mac {
description
Huang, et al. Expires March 08, 2014 [Page 64]
Internet-Draft yang-spf September 2013
"Used with address and mask couple to
express network.";
leaf source-address {
type yang:mac-address;
mandatory true;
description "A source MAC address.";
}
leaf source-address-mask {
type yang:mac-address;
mandatory true;
description "A source MAC address mask.";
}
}
leaf source-any {
type empty;
description "To express Any network or address";
}
case source-host {
description
"Use the host address combination as an
abbreviation for an address and wildcard
of address 0.0.0.0";
uses MAC-SOURCE-HOST;
}
}
}
grouping MAC-DESTINATION-NETWORK {
description
"MAC address and mask pair for destination.";
grouping MAC-DESTINATION-HOST {
description
"Choice within a case not allowed so need
this grouping.";
choice dest-address-or-name {
mandatory true;
leaf dest-host-address {
type inet:ip-address;
description
"Use the host address combination as an
abbreviation for an address and wildcard
of address 0.0.0.0";
}
leaf dest-host-name {
if-feature spf:host-by-name;
type inet:domain-name;
}
Huang, et al. Expires March 08, 2014 [Page 65]
Internet-Draft yang-spf September 2013
}
}
choice dest-network {
mandatory true;
case dest-mac {
description
"Used with address and mask couple to
express network.";
leaf dest-address {
type yang:mac-address;
mandatory true;
description "A source MAC address.";
}
leaf dest-address-mask {
type yang:mac-address;
mandatory true;
description "A source MAC address mask.";
}
}
leaf dest-any {
type empty;
description "To express Any network or address";
}
case dest-host {
description
"Use the host address combination as an
abbreviation for an address and wildcard
of address 0.0.0.0";
uses MAC-DESTINATION-HOST;
}
}
}
/* Layer 2 SPF */
augment "/spf:spfs/spf:spf" {
when "spf:spf-type = 'mac-spf'";
description
"Layer 2 Packet Filter Entry (PFE). The mac-pfes
container contains a list of mac-pfe. Each mac-pfe is
comprised of a name, an optional remark
and a rule.
A rule is referred to as 'packet-filter', although it
contains both a filter and an action.
The packet-filter requires a mandatory action (permit/deny)
and one or more options such as source-address with mask,
ethertype, vlan etc.";
Huang, et al. Expires March 08, 2014 [Page 66]
Internet-Draft yang-spf September 2013
container mac-pfes {
list mac-pfe {
key name;
ordered-by user;
leaf name {
type spf:spf-name-string;
description "Unique PFE identifier";
}
choice remark-or-mac-pfe {
leaf remark {
type spf:spf-remark;
// mandatory true;
}
case mac-pfe {
container filters {
uses MAC-SOURCE-NETWORK;
uses MAC-DESTINATION-NETWORK;
leaf ethertype {
type c-types:ether-type;
description "Ether-Type (also known as
protocol) in hex 0x0-0xffff";
}
leaf ethertype-mask {
if-feature ethertype-mask;
when "boolean(../ethertype)";
type c-types:ether-type;
default "0x0000";
description
"Ether-type mask in hex 0x0-0xFFFF.
0x0 is exactly match of the Ethertype..";
}
leaf cos {
type c-types:cos;
description "CoS value <0-7>";
}
leaf time-range {
type spf:time-range-ref;
description
"Enable packet capture on this
filter for a specify time range
by name.";
}
Huang, et al. Expires March 08, 2014 [Page 67]
Internet-Draft yang-spf September 2013
leaf vlan {
type c-types:vlan-identifier;
description "VLAN number";
}
uses spf:FILTER-COMMON;
} // container filters
uses spf:PFE-COMMON;
} // case mac-pfe
} // choice remark-or-pfe
} // list mac-pfe
} // container mac-pfes
} // augment
}
<CODE ENDS>
13. SPF-ARP Configuration YANG Module
<CODE BEGINS> file "spf-arp@2013-09-03.yang"
module spf-arp {
namespace "urn:cisco:params:xml:ns:yang:spf-arp";
// replace with IANA namespace when assigned
prefix spf-arp;
import stateless-pf { prefix spf; }
import spf-ip { prefix spf-ip; }
import spf-mac { prefix spf-mac; }
organization
"IETF NETMOD (NETCONF Data Modeling Language) Working Group";
contact
"WG Web: http://tools.ietf.org/wg/netmod/
WG List: netmod@ietf.org
WG Chair: David Kessens
david.kessens@nsn.com
WG Chair: Juergen Schoenwaelder
Huang, et al. Expires March 08, 2014 [Page 68]
Internet-Draft yang-spf September 2013
j.schoenwaelder@jacobs-university.de
Editor: Lisa Huang
yihuan@cisco.com
Editor: Alexander Clemm
alex@cisco.com
Editor: Andy Bierman
andy@yumaworks.com";
description
"This YANG module augments the 'stateless-pf' module with
configuration and operational data for ARP stateless
packet filter.
An Stateless Packet Filter (SPF), also know as an Access
Control List (SPF), is an ordered set of rules and
actions used to filter traffic.
Each set of rules and actions is represented as a Packet Filter
Entry (PFE), also know as an Access
Control Entries (PFE). Each PFE is evaluated sequentially.
When the rule matches then action for that rule is applied
to the packet.
ARP SPFs - The device applies ARP SPFs to IP traffic.
Terms and Acronyms
PFE (pfe): Packet Filter Entry
SPF (spf): Stateless Packet Filter
ARP (arp): Address Resolution Protocol
IP (ip): Internet Protocol
MAC: Media Access Control
VLAN (vlan): Virtual Local Area Network
";
revision 2013-09-03 {
description "Initial revision. ";
}
/* Identities */
identity arp-spf {
base "spf:spf-type";
Huang, et al. Expires March 08, 2014 [Page 69]
Internet-Draft yang-spf September 2013
description "ARP SPF type";
}
/* Data Nodes */
augment "/spf:spfs/spf:spf" {
when "spf:spf-type = 'arp-spf'";
description "ARP Packet FIlter Entry (PFE).";
container arp-pfes {
list arp-pfe {
key "name";
ordered-by user;
leaf name {
type spf:spf-name-string;
}
choice remark-or-arp-pfe {
leaf remark {
type spf:spf-remark;
// mandatory true;
}
case arp-pfe {
container filters {
leaf direction {
default "bi-direction";
type enumeration {
enum bi-direction;
enum request;
enum response;
}
description "ARP request/response.";
}
uses spf-ip:IP-SOURCE-NETWORK;
uses spf-ip:IP-DESTINATION-NETWORK {
when "../direction = 'response'";
}
uses spf-mac:MAC-SOURCE-NETWORK;
uses spf-mac:MAC-DESTINATION-NETWORK {
when "../direction = 'response'";
}
uses spf:FILTER-COMMON;
} // container filters
Huang, et al. Expires March 08, 2014 [Page 70]
Internet-Draft yang-spf September 2013
uses spf:PFE-COMMON;
} // case arp-pfe
} // choice remark-or-arp-pfe
} // list arp-pfe
} // container arp-pfes
} // augment
}
<CODE ENDS>
14. COMMON-TYPES YANG Module
<CODE BEGINS> file "common-types@2012-10-12.yang"
module common-types {
namespace "urn:cisco:params:xml:ns:yang:common-types";
// replace with IANA namespace when assigned
prefix c-types;
organization
"IETF NETMOD (NETCONF Data Modeling Language) Working Group";
contact
"WG Web: http://tools.ietf.org/wg/netmod/
WG List: netmod@ietf.org
WG Chair: David Kessens
david.kessens@nsn.com
WG Chair: Juergen Schoenwaelder
j.schoenwaelder@jacobs-university.de
Editor: Lisa Huang
yihuan@cisco.com
Editor: Alexander Clemm
alex@cisco.com
Editor: Andy Bierman
andy@yumaworks.com";
description
"This module contains a collection of generally useful
YANG types could be referred from multiple speciality
Huang, et al. Expires March 08, 2014 [Page 71]
Internet-Draft yang-spf September 2013
components.
Terms and Acronyms
CoS (cos): Class of Service
ICMP (icmp): Internet Control Message Protocol
IGMP (igmp): Internet Group Management Protocol
IP (ip): Internet Protocol
IPv4 (ipv4):Internet Protocol Version 4
IPv6 (ipv6): Internet Protocol Version 6
TCP (tcp): Transmission Control Protocol
ToS (tos): Type of Service
TTL (ttl): Time to Live
UDP (udp): User Datagram Protocol
VLAN (vlan): Virtual Local Area Network
";
revision 2012-10-12 {
description "Initial revision. ";
}
/* Typedefs */
typedef cos {
type uint8 {
range "0..7";
}
description
"Class of Service.
An integer that is in the range of the layer 2 CoS values.
This corresponds to the 802.1p and ISL CoS values.";
reference "IEEE 802.1p";
}
typedef tos {
type uint8 {
range "0..15";
}
description
Huang, et al. Expires March 08, 2014 [Page 72]
Internet-Draft yang-spf September 2013
"tos stands for Type of service .
The tos field are five bits in the IPv4 header.
It could specify a datagrams priority and
request a route for low-delay, high-throughput,
or highly-reliable service.
Based on these TOS values, a packet would be placed in
an prioritized outgoing queue, or take a route with
appropriate latency, throughput, or reliability.
The following are TOS field values (expressed as
binary numbers):
1000 -- minimize delay
0100 -- maximize throughput
0010 -- maximize reliability
0001 -- minimize monetary cost
0000 -- normal service
.";
reference
"RFC 791 Internet Protocol
Protocol Specification
RFC 1122 Requirements for Internet Hosts --
Communication Layers
RFC 1349 Type of Service in the Internet Protocol
Suite
RFC 2474 Definition of the Differentiated Services
Field (DS Field)
in the IPv4 and IPv6 Headers
RFC 3168 The Addition of Explicit Congestion
Notification (ECN) to IP
";
}
typedef precedence {
type uint8 {
range "0..7";
}
description
"Indicates the IP precedence.
Precedence is three bits in IP header.
Value Description
-------------------
000 (0) Routine or Best Effort
001 (1) Priority
010 (2) Immediate
Huang, et al. Expires March 08, 2014 [Page 73]
Internet-Draft yang-spf September 2013
011 (3) Flash - mainly used for Voice Signaling
or for Video.
100 (4) Flash Override
101 (5) Critical -mainly used for Voice RTP.
110 (6) Internet
111 (7) Network";
reference
"RFC 791 Internet Protocol Chapter 3.1
Protocol Specification";
}
typedef tcp-flag-type {
type bits {
bit fin {
position 0;
description "No more data from sender";
}
bit syn {
position 1;
description "Synchronize sequence numbers";
}
bit rst {
position 2;
description "Reset the connection";
}
bit psh {
position 3;
description "Push Function";
}
bit ack {
position 4;
description "Acknowledgment field significant";
}
bit urg {
position 5;
description "Urgent Pointer field significant";
}
}
description "TCP flag type";
reference "RFC 793 TRANSMISSION CONTROL PROTOCOL";
}
typedef ether-type {
type string {
pattern '0x[0-9a-fA-F]{4}';
}
description
Huang, et al. Expires March 08, 2014 [Page 74]
Internet-Draft yang-spf September 2013
"ether-type is 0x0-0xffff. The protocol number
is a four-byte hexadecimal number prefixed with 0x.
Valid protocol numbers are from 0x0 to 0xffff.
This list shows the EtherType values and their
corresponding protocol keywords:
0x0600 xns-idp Xerox XNS IDP
0x0BAD vines-ip Banyan VINES IP
0x0baf vines-echo Banyan VINES Echo
0x6000 etype-6000 DEC unassigned, experimental
0x6001 mop-dump DEC Maintenance Operation Protocol
(MOP) Dump/Load Assistance
0x6002 mop-console DEC MOP Remote Console
0x6003 decnet-iv DEC DECnet Phase IV Route
0x6004 lat DEC Local Area Transport (LAT)
0x6005 diagnostic DEC DECnet Diagnostics
0x6007 lavc-sca DEC Local-Area VAX Cluster (LAVC), SCA
0x6008 amber DEC AMBER
0x6009 mumps DEC MUMPS
0x0800 ip Malformed, invalid, or deliberately corrupt
IP frames
0x8038 dec-spanning DEC LANBridge Management
0x8039 dsm DEC DSM/DDP
0x8040 netbios DEC PATHWORKS DECnet NETBIOS Emulation
0x8041 msdos DEC Local Area System Transport
0x8042 etype-8042 DEC unassigned
0x809B appletalk Kinetics EtherTalk (AppleTalk over
Ethernet)
Huang, et al. Expires March 08, 2014 [Page 75]
Internet-Draft yang-spf September 2013
0x80F3 aarp Kinetics AppleTalk Address Resolution
Protocol (AARP)
bpdu-sap BPDU SAP encapsulated packets
bpdu-snap BPDU SNAP encapsulated packets
ipx-arpa IPX Advanced Research Projects Agency
(ARPA)
ipx-non-arpa IPX non arpa
lacp Link Aggregation Control Protocol(LACP)
encapsulated packets
pagp Port Aggregation Protocol(PAGP)
encapsulated packets
vtp VTP packets
";
}
typedef ip-protocol {
type uint8{
range "0..255";
}
description
"The Internet Protocol (IP) is the principal communications
protocol used for relaying datagrams (also known as network
packets) across an internetwork using the Internet Protocol
Suite.
IP protocol number value is 0 to 255. It is an 8 bit field
in the packet header";
reference
"IANA Protocol Numbers
RFC5237 IANA Allocation Guidelines for the Protocol Field";
}
typedef igmp-code {
//TODO: need more work. In NxOs, range is 0..15.
// Could not match the IGMP with 0..15
type uint8 ;/* {
range "0..15";
}*/
//IGMP v1 4 bits 0-15
//IGMP v2 8bits. 0-
//NXOS only support v1, but XR support v2.
//
description
"Many of these IGMP types have a 'code' field. Here is
the list of the types again with their assigned
code fields.
Huang, et al. Expires March 08, 2014 [Page 76]
Internet-Draft yang-spf September 2013
Type Name Reference
--------- ------------------------------------ ---------
0x11 IGMP Membership Query [RFC1112]
0x12 IGMPv1 Membership Report [RFC1112]
0x13 DVMRP [RFCDVMRP]
0x14 PIM version 1 [PIMv1]
0x15 Cisco Trace Messages
0x16 IGMPv2 Membership Report [RFC2236]
0x17 IGMPv2 Leave Group [RFC2236]
0x1e Multicast Traceroute Response [Fenner]
0x1f Multicast Traceroute [Fenner]
0x22 IGMPv3 Membership Report [RFC3376]
";
reference
"IANA Internet Group Management Protocol (IGMP) Type
Numbers";
}
typedef icmp-type {
type uint32 {
range "0..255";
}
description
"icmp-type is the Internet Control Message Protocol (ICMP)
'type' field.
The ICMP header starts after the IPv4 header. All ICMP
packets will have an 8-byte header and variable-sized
data section.
The first 4 bytes of the header will be consistent.
The first byte is for the ICMP type. The second byte is
for the ICMP code.
ICMP type is specified below
Type Name Reference
---- ------------------------- ---------
0 Echo Reply [RFC792]
1 Unassigned [JBP]
2 Unassigned [JBP]
3 Destination Unreachable [RFC792]
4 Source Quench [RFC792]
5 Redirect [RFC792]
6 Alternate Host Address [JBP]
7 Unassigned [JBP]
8 Echo [RFC792]
9 Router Advertisement [RFC1256]
10 Router Selection [RFC1256]
11 Time Exceeded [RFC792]
12 Parameter Problem [RFC792]
Huang, et al. Expires March 08, 2014 [Page 77]
Internet-Draft yang-spf September 2013
13 Timestamp [RFC792]
14 Timestamp Reply [RFC792]
15 Information Request [RFC792]
16 Information Reply [RFC792]
17 Address Mask Request [RFC950]
18 Address Mask Reply [RFC950]
19 Reserved (for Security) [Solo]
20-29 Reserved (for Robustness Experiment) [ZSu]
30 Traceroute [RFC1393]
31 Datagram Conversion Error [RFC1475]
32 Mobile Host Redirect [David Johnson]
33 IPv6 Where-Are-You [Bill Simpson]
34 IPv6 I-Am-Here [Bill Simpson]
35 Mobile Registration Request [Bill Simpson]
36 Mobile Registration Reply [Bill Simpson]
37-255 Reserved [JBP]";
reference
"RFC1700 ASSIGNED NUMBERS
RFC792 Internet Control Message Protocol
RFC4443 Internet Control Message Protocol (ICMPv6)
for the Internet Protocol Version 6 (IPv6)
Specification
RFC2780 IANA Allocation Guidelines For Values In
the Internet Protocol and Related Headers";
}
typedef icmp-code {
type uint32 {
range "0..255";
}
description
"ICMP subtype to the given type.
The ICMP header starts after the IPv4 header. All ICMP
packets will have an 8-byte header and variable-sized
data section.
The first 4 bytes of the header will be consistent.
The first byte is for the ICMP type. The second byte
is for the ICMP code. ";
reference "RFC2 INTERNET CONTROL MESSAGE PROTOCOL";
}
typedef vlan-identifier {
type uint16 {
range "1 .. 4095";
}
description
"This type denotes a VLAN tag. ";
reference
Huang, et al. Expires March 08, 2014 [Page 78]
Internet-Draft yang-spf September 2013
"RFC3069 VLAN Aggregation for Efficient IP Address
Allocation
IEEE 802.1Q";
}
typedef time-to-live {
type uint8 {
range "0..255";
}
description "The TTL is an 8-bit field in IP header.
The maximum TTL value is 255.";
}
}
<CODE ENDS>
15. Security Considerations
.
16. Open items from the previous revision
1. Are there any compatibility issues related to PFE ordering
because a YANG user-order list is used instead of sequence IDs?
This item is closely related to bullet item 3, see below.
2. Is an administrative function to test a packet against a
specified SPF needed? The server would return an indication of
permit or deny, and a leaf-list of the PFE entries that were
evaluated. We believe that this addition would be valuable and
have incorporated this suggestion into the "Additional
Considerations" section. We expect to move it into the data model
in the next revision.
3.Is the model applicable to multiple implementations - can other
SPF models be accommodated? We have followed up with Juniper Yang
experts, Kent Watsen and Phil Shafer, to review and check for
applicability to Junos implementation. The initial feedback from
Phil indicates that there do not seem to be any showstoppers and
that the model does seem to be applicable. However, he suggested
further scrutiny should occur. Kent identified additional Juniper
experts to scrutinize the model more closely; so far no further
comments have been received. We also followed up regarding
whether there are other standardized models of SPFs, for example
in conjunction with the Desktop Management Task Force's (DMTF) CIM
(Common Information Model). SPF is not covered by the
standardized portion of CIM, but there are vendor-specific
Huang, et al. Expires March 08, 2014 [Page 79]
Internet-Draft yang-spf September 2013
extensions by vendors. We inspected one such vendor specific
model and found that in essence the same design patterns were used
as in the model specified in this Internet Draft, with an SPF
corresponding to an ordered list of rules with filters or matching
criteria, and actions to be taken in response. It appears that
mappings between the models can be accommodated in a
straightforward manner.
17. Acknowledgements
We wish to acknowledge the helpful contributions, comments, and
suggestions that were received from Louis Fourie, Dana Blair, Tula
Kraiser, Patrick Gili, George Serpa, Martin Bjorklund, Kent Watsen,
and Phil Shafer.
18. References
18.1. Normative References
[RFC6020] Bjorklund, M., "YANG - A Data Modeling Language for the
Network Configuration Protocol (NETCONF)", RFC 6020,
October 2010.
[RFC6021] Schoenwaelder, J., "Common YANG Data Types", RFC 6021,
October 2010.
18.2. Informative References
[if-config]
Bjorklund, M., "A YANG Data Model for Interface
Management", I-D draft-ietf-netmod-interfaces-cfg-12, July
2013.
Authors' Addresses
Lisa Huang
Cisco Systems
EMail: yihuan@cisco.com
Alexander Clemm
Cisco Systems
EMail: alex@cisco.com
Huang, et al. Expires March 08, 2014 [Page 80]
Internet-Draft yang-spf September 2013
Andy Bierman
YumaWorks
EMail: andy@yumaworks.com
Huang, et al. Expires March 08, 2014 [Page 81]
