Description
Title of Invention: COMMUNICATION SYSTEM, CONTROL
DEVICE, COMMUNICATION METHOD, AND PROGRAM
Technical Field
[000 1] (REFERENCE TO RELATED APPLICATION)
The present invention is based upon and claims the benefit of the priority of Japanese
patent application No. 2010-268401, filed on December 1, 2010, the disclosure of
which is incorporated herein in its entirety by reference thereto.
The present invention relates to a communication system, a communication device, a
control device, a control method and program for a forwarding route of a packet flow,
and in particular, relates to a communication system, a control device, a commu
nication method, and a program, that realize communication by using forwarding
nodes that process a received packet, in accordance with a processing rule matching
the received packet.
Background Art
[0002] In recent years, technology known as OpenFlow has been proposed (refer to Patent
Literature 1, and Non Patent Literatures 1 and 2). In OpenFlow, communication is
taken as end-to-end flow, and routing control, recovery from failure, load balancing,
and optimization are performed in flow units. An OpenFlow switch functioning as a
forwarding node is provided with a secure channel for communication with an
OpenFlow controller that is positioned as a control device, and operates according to a
flow table in which appropriate addition or rewriting is instructed by the OpenFlow
controller. In the flow table are definitions of sets of matching rules (header fields) that
refer to packet headers, flow statistical information (Counters), and actions (Actions)
defining processing content, for each flow.
[0003] An example of a communication system using OpenFlow is described in Fig. 11. As
shown in Fig. 11, the communication system 1A is formed of a control device 110 cor
responding to the OpenFlow controller, and flow switches 121 to 126 corresponding to
OpenFlow switches. The communication system 1A can realize communication
between connected communication terminals 131 to 134.
[0004] Referring to a sequence of Fig. 12, a description is given of operation of the commu
nication system using the abovementioned OpenFlow. Here, the description is given
using an example in which the communication terminal 131 of Fig. 11 starts 2 types of
communication with the communication terminal 134.
[0005] On receiving a packet for a first communication transmitted by the communication
terminal 131 (S901 in Fig. 12), the flow switch 121 searches for a processing rule entry
having a matching rule that matches header information of the received packet, from
the flow table. As a result of the search, in a case where an entry matching the received
packet is not found, the flow switch 121, after buffering the received packet, transmits
a new flow detection notification (Packet-In) message having information of the
received packet to the control device 110 via a secure channel, and requests deter
mination of a packet forwarding route based on source and destination of the received
packet (S902 in Fig. 12).
[0006] The control device 110 that receives the request identifies an arrival destination of
the packet from information of the new flow detection notification as the commu
nication terminal 134. The control device 110 computes a route reaching the flow
switch 126 to which the identified communication terminal 134 is connected, from the
flow switch 121, as a forwarding route of the packet for the first communication. Here,
a route of flow switch 121 -> flow switch 123 -> flow switch 126 is computed as the
forwarding route of the packet for the first communication.
[0007] The control device 110 transmits a setting (FlowMod) message for a processing rule
determining a forwarding rule of a packet realizing the abovementioned forwarding
route, to flow switches in the abovementioned route, and stores processing rule entries
in respective flow tables (S903-1 to S903-3 in Fig. 12). The flow switch 121 transmits
the buffered packet using the processing rule entry that has been set (S904 in Fig. 12).
Since processing rule entries are already set in flow tables of the flow switches 123 and
126 in the packet forwarding route in the abovementioned steps S903-2 and S903-3,
the flow switches 123 and 126 that receive the packet for the first communication
transmitted from the flow switch 121 perform packet forwarding without making a
request to determine a packet route to the control device 110.
[0008] Next, on receiving a packet for a second communication transmitted by the commu
nication terminal 131 (S905 in Fig. 12), the flow switch 121 searches for a processing
rule entry having a matching rule that matches header information of the received
packet, from the flow table. Since the second communication uses a different port
number from the first, due to an entry matching the received packet not being found,
similar to the first communication, the flow switch 121, after buffering the received
packet, transmits a new flow detection notification (Packet-In) message having in
formation of the received packet to the control device 110 via the secure channel, and
requests determination of a packet forwarding route based on source and destination of
the received packet (S906 in Fig. 12).
[0009] The control device 110 that receives the request identifies an arrival destination of
the packet from information of the new flow detection notification as the commu
nication terminal 134. The control device 110 computes a route reaching the flow
switch 126 to which the identified communication terminal 134 is connected, from the
flow switch 121, as a forwarding route of the packet for the second communication.
Here, a route of the flow switch 121 -> flow switch 124 -> flow switch 126 is computed
as the forwarding route of the packet for the first communication.
[0010] The control device 110 transmits a setting (FlowMod) message for a processing rule
determining a forwarding rule of a packet realizing the abovementioned forwarding
route, to flow switches in the abovementioned route, and stores processing rule entries
in respective flow tables (S907-1 to S907-3 in Fig. 12). The flow switch 121 transmits
a buffered packet using the processing rule entry that has been set (S908 in Fig. 12).
Since processing rule entries are already set in flow tables of the flow switches 124 and
126 in the packet forwarding route by the abovementioned steps S907-2 and S907-3,
the flow switches 124 and 126 that receive the packet for the first communication
transmitted from the flow switch 121 perform packet forwarding without making a
request to determine a packet route to the control device 110.
Citation List
Patent Literature
[001 1] PTL 1: International Publication WO 2008/095010
Non Patent Literature
[0012] NPL 1: Nick McKeown, and 7 others, "OpenFlow: Enabling Innovation in Campus
Networks", [online], [search conducted September 17, 2010] Internet
URL:http://www.openflowswitch.org//documents/openflow-wp-latest.pdf
NPL 2: "OpenFlow: Switch Specification" Version 1.0.0. (Wire Protocol 0x01),
[search conducted September 17, 2010] Internet
URL:http://www.openflowswitch.org/documents/openflow-spec-vl.O.O.pdf
Summary of Invention
Technical Problem
[0013] The entire disclosures of the above cited Patent Literature and Non Patent Literatures
are incorporated herein by reference thereto.
The following analysis is given by the present invention.
As described above, according to a communication system using OpenFlow, by con
trolling a processing rule entry of a switch in a route in flow units, even with commu
nication performed between the same communication terminals, if flow features are
different (port numbers in examples of Fig. 11 and Fig. 12 are different), it is possible
to realize flexible routing control, such as it being possible to select different routes.
[0014] On the other hand, however, the communication system using OpenFlow performs
routing control of all flow switches within the communication system by a control
device such as an OpenFlow controller. As a result, each time a new flow is generated,
there is a problem in that the control device has to set a processing rule in all flow
switches in a route taken by a new flow, and control load increases.
[0015] The present invention has been made in view of the abovementioned matter and it is
an object to provide a communication system, a control device, a communication
method, and a program that can lessen the load on the control device that centrally
controls forwarding nodes thereunder, as in the OpenFlow controller described above.
Solution to Problem
[0016] According to a first aspect of the present invention, there is provided a commu
nication system that comprises: a plurality of forwarding nodes for which is set a
processing rule(s) for forwarding a received packet in accordance with a route set in
advance, with a processing rule being selected based on a path identifier included in
the received packet, to perform packet forwarding; and a control device that sets the
processing rule beforehand in the forwarding node(s) in the route, and causes a
forwarding node positioned at a start point of the route to add a path identifier in ac
cordance with the route to a received packet and causes a forwarding node positioned
at an end point of the route to delete the path identifier from a received packet, on a
predetermined occasion.
[0017] According to a second aspect of the present invention, there is provided a control
device, connected to a plurality of forwarding nodes for which is set a processing
rule(s) that forwards a received packet in accordance with a route set in advance, with
a processing rule being selected based on a path identifier included in the received
packet, to perform packet forwarding; the control device comprising: a first unit setting
the processing rule beforehand in the forwarding nodes in the route; and a second unit
causing a forwarding node positioned at a start point of the route to add a path
identifier in accordance with the route to a received packet and causing a forwarding
node positioned at an end point of the route to delete the path identifier from a received
packet, on a predetermined occasion.
[0018] According to a third aspect of the present invention, there is provided a commu
nication method that comprises: a step in which, when a new flow is generated, a
forwarding node positioned at a start point of a route that has been computed in
advance adds a path identifier to a received packet, based on an instruction from a
control device, and then forwards to a subsequent hop;
a step in which a forwarding node that is in a route computed in advance selects a
processing rule based on the path identifier included in a received packet, from among
processing rules set in advance by the control device, and forwards the received
packet; and
a step in which the forwarding node positioned at an end point of the route restores
the received packet to before the path identifier was added, based on an instruction
from the control device, and then forwards to a subsequent hop. The present method is
tied up with specific apparatuses, known as a transfer node and a control device that
configure the communication system.
[0019] According to a fourth aspect of the present invention, there is provided a program
executed in a control device connected to a plurality of forwarding node for which is
set a processing rule that forwards a received packet in accordance with a route set in
advance, with a processing rule being selected based on a path identifier included in
the received packet, and packet forwarding being performed; wherein the program
executes a process of setting the processing rule beforehand in the forwarding node in
the route, and a process, at a predetermined occasion, of respectively executing
addition and deletion of the path identifier in accordance with the route, in forwarding
nodes positioned at a start point and an end point of the route. It is to be noted that that
the program can be recorded in a computer readable storage medium, which may be a
non-transient recording medium. That is, the present invention can be embodied as a
computer program product.
Advantageous Effects of Invention
[0020] According to the present invention it is possible to lighten the load on a control
device that centrally controls forwarding nodes thereunder.
Brief Description of Drawings
[0021] [fig. l]Fig. 1 is a diagram for describing an outline of the present disclosure;
[fig.2]Fig. 2 is a diagram showing a configuration example of a communication system
according to a first exemplary embodiment of the present disclosure;
[fig.3]Fig. 3 is a diagram showing a configuration example of a control device
according to the first exemplary embodiment;
[fig.4]Fig. 4 is a diagram for describing content stored in a path information storage
unit of the control device of the first exemplary embodiment;
[fig.5]Fig. 5 is a sequence diagram showing an operational example (pre-setting of
processing rule) of the first exemplary embodiment;
[fig.6]Fig. 6 is a flow diagram representing processing flow of the control device after
step SO13 of Fig. 5.
[fig.7]Fig. 7 is a sequence diagram showing an operational example (when a new flow
is generated) of the first exemplary embodiment;
[fig.8]Fig. 8 is a diagram showing a configuration example of a communication system
according to a second exemplary embodiment of the present disclosure;
[fig.9]Fig. 9 is a flow diagram representing processing flow of a control device
according to the second exemplary embodiment;
[fig. 10] Fig. 10 is a diagram for describing content stored in a path information storage
unit of the control device of the second exemplary embodiment;
[fig. 1l]Fig. 11 is a diagram for describing a configuration example of a commu
nication system of Non Patent Literatures 1 and 2; and
[fig. 12] Fig. 12 is a sequence diagram for describing an operational example of the
communication system of Non Patent Literatures 1 and 2.
Description of Embodiments
[0022] First, an outline of the present disclosure is described. As shown in Fig. 1, the present
invention can be implemented by forwarding nodes 2 1 to 26 wherein a processing rule
that forwards a received packet in accordance with routes set in advance (pre-set routes
A and B of Fig. 1) is set, with a processing rule being selected based on a path
identifier included in the received packet, and packet forwarding being performed; and
a control device 10 that sets the processing rule beforehand in the forwarding nodes
(forwarding nodes 21, 23, 24, and 26 in Fig. 1) that are in the route. The control device
10 also causes a forwarding node positioned at a start point of the route to add a path
identifier in accordance with the route to a received packet and causes a forwarding
node positioned at an end point of the route to delete the path identifier from a received
packet when a new flow is generated (forwarding nodes 2 1 and 26 in Fig. 1). It is to be
noted that drawing reference symbols noted in this outline are added for convenience
to respective elements as one example in order to aid understanding, and are not
intended to limit the invention to modes of the drawings shown.
[0023] Specifically, implementation is realized by a step in which, when a new flow is
generated from a communication terminal 131 with a communication terminal 134 as a
destination, the forwarding node 2 1 positioned at a start point of a route has been
computed in advance, adds a path identifier (for example, a path identifier A for a pre
set route A) to the received packet, based on an instruction from the control device,
and then forwards to a subsequent hop; a step in which the forwarding node 24 that is
in the route computed in advance selects a processing rule based on the path identifier
(for example, path identifier A) included in the received packet, and forwards the
received packet to the forwarding node 26; and a step in which the forwarding node 26
positioned at the end point of the route restores the received packet to before the path
identifier was added, based on an instruction from the control device, and then
forwards to a subsequent hop.
[0024] In the same way, it is possible to select the pre-set route B as a forwarding route of
the new flow, and to perform packet forwarding.
[0025] As described above, according to the present invention, since there is no need for a
request to set a processing rule from a forwarding node (excepting forwarding nodes at
the start and end points) in a route that has been set in advance, it is possible to reduce
the load on the control device 10.
[0026] (First Exemplary Embodiment)
Next, a detailed description is given concerning a first exemplary embodiment of the
present disclosure making reference to the drawings. Fig. 2 is a diagram showing a
configuration example of a communication system according to the first exemplary
embodiment of the present disclosure. Referring to Fig. 2, the communication system 1
formed from a control device 10 and a forwarding node group 20 is shown.
[0027] The forwarding node group 20 is formed from a plurality of forwarding nodes 2 1 to
26.
[0028] In addition, communication terminals 131 to 134 are connected to the commu
nication system 1.
[0029] It is to be noted that symbols such as # 1 and #2 noted beside the forwarding nodes
indicate port numbers with respect to the respective forwarding nodes.
[0030] In order to simply describe the present exemplary embodiment, the two terms of
"edge node" and "core node" are defined here. The edge node indicates a forwarding
node at a start point or an end point of a path to be described later. A core node
indicates a forwarding node that is not a start point nor an end point of a path to be
described later.
[0031] Here, in order to efficiently describe the exemplary embodiment below, the words
"route" and "path" are defined. A "route" relates to information formed by one or more
forwarding nodes that are passed by data packets transmitted and received between
communication terminals performing communication via the communication system 1.
Furthermore a "route" possesses directionality. For example, when two communication
terminals use the same forwarding nodes for going and returning, routes for going and
returning are different. In contrast to this, a "path" is built from one or more "routes",
and relates to setting granularity of a packet forwarding rule in a forwarding node
belonging to core nodes in the communication system 1. As a result, viewed from a
specific route, an edge node and a core node exist exclusively from one another, but
from a viewpoint of a forwarding node, forwarding nodes also exist that take roles of
both an edge node and a core node, being an edge node in a certain route and a core
node in another route.
[0032] Furthermore, in the following exemplary embodiment, a description is given that
assumes packet forwarding is performed such that, with an edge node, a path identifier
described later for identifying a path is embedded in a destination MAC address of a
packet, and a core node looks at the path identifier embedded in the destination MAC
address to decide an action. A place where the path identifier is embedded can be a
wide variety of places such as an unused header field or the like.
[0033] Fig. 3 is a diagram representing a detailed configuration of the control device 10 of
Fig. 2. Referring to Fig. 3, the control device 10 is configured by being provided with a
communication unit 11 that performs communication with the forwarding nodes 2 1 to
26, a control message processing unit 12, an action computation unit 13, a forwarding
node management unit 14, a topology management unit 15, a terminal location in
formation management unit 16, a processing rule management unit 17A, a processing
rule storage unit 17B, a path information management unit 18A and a path information
storage unit 18B. These respectively operate as follows.
[0034] The control message processing unit 12 analyzes a control message received from the
forwarding nodes 2 1 to 26 and delivers control message information to a relevant
processing means inside the control device 10.
[0035] The action computation unit 13 calls for an action to be executed in a forwarding
node in the forwarding route of a packet based on location information of a commu
nication terminal managed by the terminal location information management unit 16,
and path information built in the path information management unit 18A.
[0036] The forwarding node management unit 14 manages capability (for example, number
and type of ports, types of action supported, and the like) of forwarding nodes
controlled by the control device 10.
[0037] The topology management unit 15 builds network topology information based on
connection relationships of forwarding nodes collected via the communication unit 11.
Furthermore, if a change is seen in the topology information, notification thereof is
provided to the path information management unit 18A.
[0038] The terminal location information management unit 16 manages information for
identifying the location of a communication terminal connected to the communication
system. In the present exemplary embodiment, a description is given in which, with an
IP address as information for distinguishing the communication terminal, being in
formation for identifying the location of the communication terminal, usage is made of
information for distinguishing a forwarding node to which the communication terminal
is connected and information of a port thereof. Clearly, a terminal and its location may
also be identified using other information instead of these items of information.
[0039] The processing rule management unit 17A performs management regarding what
type of processing rule is set in which forwarding node. Specifically, a result
calculated in the action computation unit 13 is recorded in the processing rule storage
unit 17B as a processing rule, and the processing rule is set in the forwarding node, and
recording information of the processing rule storage unit 17B is updated in response to
a case also where a change occurs in the processing rule set in the forwarding node, by
a processing rule deletion notice from the forwarding node.
[0040] The path information management unit 18A manages path information used in com
munication inside the communication system 1. Specifically, when there is a change in
the topology configuration of the communication system 1, a route used in commu
nication is derived, a path having setting granularity of a processing rule set in a core
node is computed from the route, a unique identifier is allocated within the forwarding
node group 20 for each path and recorded in the path information storage unit 18B, and
in coordination with the action computation unit 13, the processing rule set in the core
node (not depending on a request from the core node) is updated.
[0041] Furthermore, the path information management unit 18A returns the relevant path in
formation when path information is requested from the action computation unit 13.
Fig. 4 is a diagram for describing path information stored in the path information
storage unit 18B. The path information is formed of a start point forwarding node, an
end point forwarding node, a path identifier, and route information.
[0042] The start point forwarding node and the end point forwarding node are forwarding
nodes respectively relating to the start point and the end point of the route. As in
formation for distinguishing the forwarding nodes used here, the IP address or MAC
address of the forwarding nodes may be used, or other dedicated identifiers may be
used.
[0043] The path identifier is information of fixed length for uniquely distinguishing, within
the forwarding node group 20, path information defined at the route information
described later.
[0044] The route information is information indicating specific information of a path
selected as a path used in communication between communication terminals by the
path information management unit 18A, by a combination of one or more of
forwarding node(s) and port information thereof. For example, with regard to route in
formation of an entry distinguished by Path #000X of Fig. 4, passage from port #2 of
the forwarding node 2 1 to port #4 of the forwarding node 23, and arrival at the
forwarding node 26 which is an end forwarding node is understood (refer to pre- set
route B of Fig. 1).
[0045] Furthermore, as is understood from a table configuration of Fig. 4, a path in the
present exemplary embodiment, in a single direction, corresponds 1 to 1 with a route
specified between 2 specific forwarding nodes.
[0046] It is to be noted that, in the abovementioned configuration, in a case where it is not
necessary to hold a processing rule in the control device 10, the processing rule storage
unit 17B can be omitted. In addition, a configuration can also be adopted in which the
processing rule storage unit 17B and the path information storage unit 18B are
provided in a separate external server or the like.
[0047] The control device 10 as described above can also be realized by a configuration in
which the abovementioned the path information management unit 18A is added, based
on the OpenFlow controller of Non Patent Literatures 1 and 2.
[0048] Furthermore, respective parts (processing means) of the control device 10 shown in
Fig. 1 can also be realized by a computer program that causes the abovementioned r e
spective processing to be executed in a computer forming the control device 10, using
hardware thereof.
[0049] It is to be noted that respective parts (processing means) of the control device 10
shown in Fig. 2 can also be realized by a computer program that causes the abovementioned
respective processing to be executed in a computer forming the control
device 10, using hardware thereof.
[0050] On receiving a packet, the forwarding nodes 2 1 to 26 find a processing rule having a
matching rule (reference rule) that matches the received packet, from the processing
rule storage unit that stores processing rules, and implement processing (for example,
forwarding to a specific port, flooding, dropping, and the like) as in an action attached
to the processing rule.
[0051] Furthermore, each time a packet is processed, the forwarding nodes 2 1 to 26 reset a
timer (providing time out information) within an action field of the processing rule in
question. When the timer becomes (counts up to) 0, the forwarding nodes 2 1 to 26
delete the processing rule in question from the processing rule storage unit. In this way,
it is possible to prevent a situation where a processing rule that is no longer used
remains forever and an unintended action is executed.
[0052] It is to be noted that the forwarding nodes 2 1 to 26 described above can also be
realized in a configuration equivalent to an OpenFlow switch of Non Patent Literatures
1 and 2. In this case, the respective processing rule storage unit corresponds to a flow
table storing flow entries of an OpenFlow switch of Non Patent Literature 1.
[0053] Next, a detailed description is given concerning operation of the present exemplary
embodiment, making reference to the drawings. First, a description is given concerning
pre-setting of a processing rule to a core node by the control device 10.
[0054] Fig. 5 is a sequence diagram representing a series of procedures when the control
device 10 pre-sets a processing rule in a core node.
[0055] Referring to Fig. 5, first, when the forwarding node 2 1 starts, a control channel is
built with the control device 10 (S001 in Fig. 5). The control channel established here
is a connection of a TCP (Transmission Control Protocol) using a port number for the
control channel, or an SSL (Secure Socket Layer). The forwarding node 2 1 gives noti
fication of information about itself to the control device 10, via a forwarding capability
notification message (S002 in Fig. 5).
[0056] Here, for information transmitted by the forwarding node 2 1 to the control device 10,
for example, an identifier of the forwarding node itself (the reference symbol 2 1 of the
forwarding node in the present exemplary embodiment), respective port information,
information of supported actions, and the like are cited. As the control device 10 and
the forwarding node 21, in a case of using the OpenFlow controller and OpenFlow
switch of Non Patent Literatures 1 and 2, it is possible to use a Feature Request/Reply
message, as the forwarding capability notification message.
[0057] Besides, as an identifier of forwarding node itself, an IP address, for example, can be
transmitted. Clearly, information other than the IP address may be transmitted as an
identifier. The port information is, for example, a port identifier of each port of a
forwarding node, types of links (link type information) connected to the respective
ports, and information related to links connected to the respective ports. As an
identifier of the abovementioned port, it is possible to use, for example, a port number.
The link type information is information representing, for example, link types of
different frame types such as Ethernet (registered trademark), fiber channel, and the
like. The content of information related to links changes according to the link type. For
example, in a case where the link type is Ethernet (registered trademark), a MAC
address, supported communication speed, communication system or method (for
example, full duplex, half duplex), and the like can be cited as information related to
links. The information related to the links can also be referred to as link attribute in
formation. The action information is, for example, in addition to normal packet
forwarding, information of supported actions such as MAC address conversion, IP
address conversion, L4 port number conversion, and the like.
[0058] On receiving a forwarding capability notification message from the forwarding node
21, the control device 10 stores content thereof in the forwarding node management
unit 14.
[0059] In the same way, the control device 10 builds the forwarding nodes 22 to 26 and the
control channel, and stores forwarding capability of the forwarding nodes 22 to 26 in
the forwarding node management unit 14 (S003 to S012 of Fig. 5).
[0060] When collection of forwarding capabilities of the respective forwarding nodes is
completed, the control device 10 performs a topology search within the forwarding
node group 20, as advance preparation in order to derive path information that a core
node uses in forwarding of data packets, and stores a result thereof in the topology
management unit 15 (S013 in Fig. 5).
[0061] When the topology information is updated, the path information management unit
18A performs derivation of path information used by the core node in forwarding a
data packet, and sets a processing rule based on the path information in forwarding
nodes (forwarding nodes 23 and 24 of Fig. 2 and Fig. 5) that form core nodes in col
laboration with the action computation unit 13 (S014 and S015 of Fig. 5).
[0062] Next, a description is given concerning flow in setting a processing rule by the path
information management unit 18A of the abovementioned control device 10, making
reference to Fig. 6.
[0063] On detecting that a topology state has changed according to a notification or the like
from the topology management unit 15 (step S101), the path information management
unit 18A computes a route used in communication between communication terminals
based on the topology information after the change (step S102).
[0064] Next, the path information management unit 18A allocates a path identifier to the
computed route, to be recorded in the path information storage unit 18B as path in
formation (step S104). When the path information management unit 18A gives noti
fication of path information to the action computation unit 13, the action computation
unit 13 creates a processing rule to be set in a core node from among received path in
formation (step S105), and sets the processing rule in the forwarding node in question
(step S106). Here, the processing rule created by the action computation unit 13 uses a
destination MAC address in which a path identifier is embedded as a matching rule
(reference rule), and an action is described that forwards to a port number specified by
route information of the path information (refer to Fig. 4).
[0065] Next, referring to Fig. 7, a description is given concerning flow of actual commu
nication using a processing rule that has been pre-set as described above. In the de
scription below, a description is given using an example in which the communication
terminal 131 of Fig. 2 starts 2 types of communication in which different ports are
used, with the communication terminal 134 (as a destination).
[0066] First, the communication terminal 131 transmits a data packet for a first commu
nication to a destination of the communication terminal 134 (S201 in Fig. 7).
[0067] On receiving the data packet, the forwarding node 2 1 searches for a processing rule
having a matching rule (reference rule) that matches the received packet, from the
processing rule storage unit 17B. However, since the packet is a first packet from the
communication terminal 131 with a destination of the communication terminal 134, a
relevant processing rule does not exist. Therefore, after buffering the received packet,
the forwarding node 2 1 transmits a new flow detection notification to the control
device 10 (S202 in Fig. 7). This new flow detection notification includes information
necessary for identification and creation of a processing rule (for example, MAC
address, IP address, port number (each including both a source and a destination)), and
packet reception port information.
[0068] It is to be noted that instead of sending only the information necessary for identi
fication and creation of a processing rule as described above to the control device 10,
the forwarding node 2 1 may transmit also the received packet itself to the control
device 10.
[0069] On receiving the new flow detection notification, the control device 10 performs
identification of a source IP address, the forwarding node that detects the new flow,
and an input port thereof, from information included in the new flow detection notification,
in the action computation unit 13. If there is a change in location information
of a communication terminal (communication terminal 131) having the same IP
address stored in the terminal location information management unit 16, the action
computation unit 13 records the forwarding node and input port information thereof
included in the new flow detection notification as location information.
[0070] Next, the action computing unit 13 refers to the terminal location information
management unit 16, and identifies location information of the communication
terminal 134, from the destination IP address included in the new flow detection noti
fication.
[0071] When the location identification of the source and destination communication
terminals is completed, the action computation unit 13 requests path information corre
sponding to a route with a start point forwarding node as the forwarding node 2 1 and
an end point forwarding node as the forwarding node 26, to the path information
management unit 18A.
[0072] The path information management unit 18A refers to the path information storage
unit 18B, selects a path identifier from information of the start point and end point
forwarding nodes, and gives notification of corresponding path information to the
action computation unit 13.
[0073] It is to be noted that in a case where there are a plurality of path identifiers sharing
the start point and end point forwarding nodes, selection may be made of one path
identifier from among them, based on a predetermined algorithm. In the present
exemplary embodiment, there is no particular limitation with regard to this algorithm,
but, citing an example, consideration can be given to an algorithm such as random
selection, round robin selection, or collaborating with the processing rule management
unit 17A to make a selection with fewer used path identifiers. Here, Path #000X is
selected as a path identifier.
[0074] On obtaining the path information, the action computation unit 13 determines a
matching rule (reference rule) for the new processing rule and computes an action to be
executed by the forwarding node 2 1 and the forwarding node 26 that are edge nodes.
Here, as the matching rule (reference rule), the IP address and port number of both
source and destination are determined respectively with respect to the communication
terminals 131 and 134.
[0075] After converting the destination MAC address to Path #000X in the forwarding node
2 1 that is the start point forwarding node, the action computation unit 13 computes an
action forwarding to port #2 of forwarding node 2 1 based on the path information. As
to the forwarding node 26 that is the end point forwarding node, after converting
(restoring) the destination MAC address to the MAC address of the communication
terminal 134, the action computation unit 13 computes an action forwarding to port #2
of forwarding node 26 that is location information for the communication terminal
134. Above conversion of MAC address to Path #000X corresponds to the addition
and deletion of the path identifier.
[0076] The control device 10 creates a processing rule based on the determined matching
rule (reference rule) and action, to be set in the forwarding nodes 2 1 and 26 that are the
start point and the end point forwarding nodes (S203-1 and S203-2 in Fig. 7).
[0077] After setting the processing rule, the control device 10 registers the processing rule
set in the forwarding nodes 2 1 and 26 in the processing rule management unit 17A.
[0078] As described above, when setting of the processing rule is complete, the forwarding
node 2 1 converts the MAC address of the first data packet, which is buffered, in ac
cordance with the processing rule, and then forwards to the forwarding node 23. Since
pre-setting of the processing rule has already been performed, in the forwarding node
23 that receives this packet, being on a route corresponding to the path information
(refer to S014 of Fig. 5), the packet is forwarded in sequence to the forwarding nodes
21, 23, and 26, and reaches the communication terminal 134 (S204 of Fig. 7).
[0079] Next, the communication terminal 131 transmits a data packet for a second commu
nication belonging to a communication flow different to the abovementioned data
packet, to a destination of the communication terminal 134 (S205 in Fig. 7).
[0080] On receiving the data packet, the forwarding node 2 1 searches for a processing rule
having a matching rule (reference rule) that matches the received packet, from the
processing rule storage unit 18B. However, since with regard to this packet, the port
number used for the first data packet is different, no relevant processing rule exists.
Therefore, after buffering the received packet, the forwarding node 2 1 transmits a new
flow detection notification to the control device 10 (S206 in Fig. 7).
[0081] Subsequent processing is the same as processing for the first data packet, outside of
selection of Path #000Y (refer to Fig. 4) by the control device 10 as the path identifier.
When setting of the processing rule in the start point and end point forwarding nodes is
completed, and a packet is transmitted in accordance with the processing rule from the
start point forwarding node 21, since pre-setting of the processing rule has already
been performed, in the forwarding node 24 that receives this packet, (refer to SO15 of
Fig. 5), the packet is forwarded in sequence to forwarding nodes 21, 24, and 26, and
reaches the communication terminal 134 (S208 of Fig. 7).
[0082] Thereafter, the packet with a destination of the communication terminal 134 from the
terminal device 131 is transmitted by 2 routes, in accordance with the 2 processing
rules described above (S209 and S210 of Fig. 7).
[0083] As described above, according to the present exemplary embodiment, since the new
flow detection notification from the core node is not received, and furthermore, it is
sufficient if the control device 10 sets the processing rule in the start point and end
point forwarding nodes, it is possible to reduce the load on the control device 10 and to
perform high speed packet forwarding.
[0084] According to the present exemplary embodiment, even with packet for which com
munication is performed between the same communication terminals, it is possible to
forward packets belonging to different communications by respectively different
routes. A reason for this is that a processing rule is set in which information, which is
held by both edge nodes and core nodes, referred to as path identifiers, is taken as a
matching rule (reference rule).
[0085] Furthermore, as is also clear from the exemplary embodiment described above, the
present disclosure can be applied without depending on the size of the communication
system. A reason for this is because of a feature in that, since the path itself is specified
by an identifier of fixed length, without describing path information by a forwarding
node group on a route or an action sequence thereof, or identifying a flow, the in
formation amount is constant and does not depend on the forwarding nodes passed
through.
[0086] Furthermore, in the present exemplary embodiment, since a system is used in which
the path identifiers are embedded in arbitrary fields that can be restored within a
packet, there is also a successful result whereby processing need not be added for
packet fragmenting and restoration thereof in a forwarding node that is a start point or
an end point of the route.
[0087] In the same way, in the present exemplary embodiment processing with respect to the
core nodes is also simplified, and there is no need for additional processing such as
label replacing as performed in MPLS (Multi-Protocol Label Switching).
[0088] It is to be noted that in the exemplary embodiment described above, when the new
flow detection notification is received, the control device 10 only sets a route in a
transmission direction of the data packet, but setting of both directions (a direction
from the communication terminal 131 to the communication terminal 134, and a
direction from the communication terminal 134 to the communication terminal 131)
may be done at the same time. In this case, the action computation unit 13 may reverse
the start point and end point forwarding nodes, and check once again with the path in
formation management unit 18A, and may also return path information of both d i
rections, with regard to a request for path information of a single direction by the path
information management unit 18A.
[0089] Furthermore, in the exemplary embodiment described above, the destination MAC
address was used as a field in which the path identifier is embedded, but there is no
particular limitation with regard to field type. For example, it is possible to use a field
that can be restored at the end point forwarding node, such as the source MAC address,
the source IP address, the destination IP address, and the like, as a field with the path
identifier embedded. In addition it is possible to employ a system in which an action
adding a new header, as in encapsulation, is set, and the path identifier is embedded in
this added header.
[0090] Furthermore, in the exemplary embodiment described above, a description was given
in which the control device 10 performs setting of a processing rule with arrival of a
data packet at a forwarding node as a trigger event, but the processing rule may also be
set at an occasion other than the data packet reception. For example, if it becomes clear
that traffic is generated at a decided point in time, by traffic information and the like
collected from forwarding nodes, a processing rule may be set in the start point and
end point forwarding nodes in advance.
[0091] (Second Exemplary Embodiment)
Next, a detailed description is given concerning a second exemplary embodiment of
the present disclosure making reference to the drawings. In the first exemplary em
bodiment described above, path information and route corresponded 1 to 1, but con
sideration can be given to the fact that, in accordance with topology becoming com
plicated, the number of path identifiers, that is, of entries of processing rules set in core
nodes, will increase. Therefore, in the second exemplary embodiment, plural routes,
which use the same forwarding node as end point forwarding nodes of the routes, are
consolidated, and one path identifier is given. Since the basic configuration otherwise
is the same as in the first exemplary embodiment described above, a description is
given below centered on points of difference.
[0092] Fig. 8 is a diagram showing a configuration example of a communication system of
the second exemplary embodiment of the present invention. A point of difference from
a configuration of the first exemplary embodiment described above resides in the point
that a communication terminal 135 is connected to a port #5 of the forwarding node 23.
[0093] In the present exemplary embodiment, pre-setting of a processing rule to a core node
by a control device 10 is performed (refer to Fig. 5). Fig. 9 is a diagram representing
flow in setting a processing rule by a path information management unit 18A of the
control device 10 of the present exemplary embodiment.
[0094] A point of difference from the first exemplary embodiment shown in Fig. 6 resides in
the point that, after finishing route computation (step SI02), consolidated processing of
route information is performed (step S103).
[0095] In the consolidated processing of the route information described above, the same
forwarding node is used as an end point forwarding node in the routes, and processing
is performed to give the same path identifier to plural routes having duplicated paths.
[0096] Fig. 10 shows a configuration example of path information stored in a path in
formation storage unit 18B. In the example of Fig. 8, similar to a path information
storage unit of the first exemplary embodiment described using Fig. 4, the path information
is configured by a path identifier, a start point forwarding node, an end point
forwarding node, and route information. However, in order to improve readability, the
arrangement of constituent elements in the table is changed (different) from respect to
Fig. 4.
[0097] In the example of Fig. 8, a path identifier Path #000X is allocated to a route with a
forwarding node 2 1 as a start point forwarding node, and a forwarding node 26 as an
end point forwarding node, encompassing a route with a forwarding node 23 as a start
point forwarding node, and the forwarding node 26 as an end point forwarding node.
Furthermore, a path identifier Path #000Y is allocated to a route tree configured by a
route with the forwarding node 2 1 as a start point forwarding node, and the forwarding
node 26 as an end point forwarding node, and a route with forwarding node 22 as a
start point forwarding node, and the forwarding node 26 as an end point forwarding
node.
[0098] Pre-setting of a processing rule thereafter and operations after a new flow detection
notification are the same as the first exemplary embodiment described above.
[0099] In the present exemplary embodiment, it is possible to obtain an effect of being able
to further reduce the number of processing rules set in a core node, by consolidating a
plurality of routes in one path.
[0100] A description has been given above of preferred exemplary embodiments of the
present disclosure, but the present invention is not limited to the abovementioned
exemplary embodiments, and further modifications, substitutions, and adjustments can
be added within a scope that does not depart from a fundamental technological concept
of the present invention. For example, the number of forwarding nodes shown in each
of the exemplary embodiments described above is an example, and the number thereof
is not limited.
[0101] Furthermore, in the exemplary embodiments described above there was only one
forwarding node group included in the communication system 1, but a plurality of
forwarding node groups is also possible. In this case, the forwarding node group is
regarded as one virtual forwarding node, and it is possible to perform management by
allocating a path identifier to each forwarding node group. In a case where commu
nication is performed via a plurality of forwarding node groups, processing for a case
where there is one forwarding node group as described in the present exemplary em
bodiment may be performed in each of the forwarding node groups.
[0102] Furthermore, in the exemplary embodiments described above, a description was
given in which packets for routing control are Ethernet (registered trademark) frames,
but an IP packet that does not include an Ethernet (registered trademark) header is also
possible. It is to be noted that each disclosure of the abovementioned Patent Literature
and Non Patent Literatures is incorporated herein by reference thereto. Modifications
and adjustments of exemplary embodiments are possible within the bounds of the
entire disclosure (including the scope of the claims) of the present invention, and also
based on fundamental technological concepts thereof. Furthermore, a wide variety of
combinations and selections of various disclosed elements is possible within the scope
of the claims of the present invention. That is, the present invention clearly includes
every type of transformation and modification that a person skilled in the art can
realize according to the entire disclosure including the scope of the claims and to tech
nological concepts thereof.
Finally, preferred modes of the present disclosure are summarized in the following.
(First Mode)
(Refer to the communication system according to the first aspect described above.)
(Second Mode)
With regard to the first mode, the communication system wherein a control device
causes the forwarding nodes positioned at a start point and an end point of said route to
execute addition and deletion of the path identifier in accordance with the route, at an
occasion when a new flow detection notification is received from the forwarding node.
(Third Mode)
With regard to the first or the second mode, the communication system wherein the
control device is provided with: a route information storage unit that manages a
plurality of routes formed by the forwarding nodes and path identifiers associated with
said routes respectively; a route information management unit that selects an ap
propriate route from the route information storage unit, when a new flow is generated;
and an action computation unit that sets a processing rule in advance, in a forwarding
node in a route selected from the route information storage unit, and also sets a
processing rule respectively executing addition and deletion of the path identifier in ac
cordance with the route, in forwarding nodes positioned at a start point and an end
point of the selected route.
(Fourth Mode)
With regard to any of the first to the third mode, in the communication system, the
control device performs re-computation of a route between arbitrary forwarding nodes
and sets a processing rule corresponding to the route, each time a change of topology
of a network formed by the forwarding nodes is detected.
(Fifth Mode)
With regard to the fourth mode, the communication system gives the same path
identifier to routes having an overlapping section, with the same end point forwarding
node, among routes between the arbitrary forwarding nodes.
(Sixth Mode)
(Refer to the control device according to the second aspect described above.)
(Seventh Mode)
With regard to the sixth mode, the control device causes the forwarding nodes p o
sitioned at a start point and an end point of said route to execute addition and deletion
of the path identifier in accordance with the route, at an occasion when a new flow
detection notification is received from the forwarding node.
(Eighth Mode)
With regard to the sixth or seventh mode, the control device is further provided with: a
route information storage unit that manages a plurality of routes formed by the
forwarding nodes and path identifiers associated with said routes respectively; a route
information management unit that selects an appropriate route from the route in
formation storage unit, when a new flow is generated; and an action computation unit
that sets a processing rule in advance, in a forwarding node in a route selected from the
route information storage unit, and also sets a processing rule respectively executing
addition and deletion of the path identifier in accordance with the route, in forwarding
nodes positioned at a start point and an end point of the selected route.
(Ninth Mode)
With regard to any of the sixth to the eighth modes, the control device performs recomputation
of a route between arbitrary forwarding nodes and sets a processing rule
corresponding to the route, each time a change of topology of a network formed by the
forwarding nodes is detected.
(Tenth Mode)
With regard to the ninth mode, the control device gives the same path identifier to
routes having an overlapping section, with the same end point forwarding node, among
routes between the arbitrary forwarding nodes.
(Eleventh Mode)
(Refer to the communication method according to the third aspect described above)
(Twelfth Mode)
(Refer to the program according to the fourth aspect described above)
Each of the function units and method steps may be comprehended as and im
plemented by means for performing each function or steps, particularly relating to the
processing means or steps for calculating particular physical parameters concerned.
Reference Signs List
I communication system
10 control device
I I communication unit
12 control message processing unit
13 action computation unit
forwarding node management unit
topology management unit
terminal location information management unit
A processing rule management unit
B processing rule storage unit
A path information management unit
B path information storage unit
forwarding node group
- 26 forwarding node
1 - 135 communication terminal
WO 2012/073409 PCT/JP2011/005106
Claims
[Claim 1] A communication system, comprising:
a plurality of forwarding nodes for which is set a processing rule(s) for
forwarding a received packet in accordance with a route set in advance,
with a processing rule being selected based on a path identifier included
in said received packet, to perform packet forwarding; and
a control device that sets said processing rule beforehand in said
forwarding node(s) in said route, and causes a forwarding node p o
sitioned at a start point of said route to add a path identifier in ac
cordance with said route to a received packet and causes a forwarding
node positioned at an end point of said route to delete said path
identifier from a received packet, on a predetermined occasion.
[Claim 2] The communication system according to claim 1, wherein a control
device causes the forwarding nodes positioned at a start point and an
end point of said route to execute the addition and deletion of a path
identifier in accordance with said route, at an occasion when a new
flow detection notification is received from said forwarding node.
[Claim 3] The communication system according to claim 1 or 2, wherein said
control device comprises:
a route information storage unit that manages a plurality of routes
formed by said forwarding nodes and path identifiers associated with
said routes respectively;
a route information management unit that selects an appropriate route
from said route information storage unit, when a new flow is generated;
and
an action computation unit that sets a processing rule in advance, in a
forwarding node in a route selected from said route information storage
unit, and also sets a processing rule respectively executing addition and
deletion of the path identifier in accordance with said route, in
forwarding nodes positioned at a start point and an end point of said
selected route.
[Claim 4] The communication system according to any one of claims 1 to 3,
wherein said control device performs re-computation of a route
between arbitrary forwarding nodes and sets a processing rule corre
sponding to said route, each time a change of topology of a network
formed by said forwarding nodes is detected.
[Claim 5] The communication system according to claim 4, wherein said commuWO
2012/073409 PCT/JP2011/005106
nication system gives the same path identifier to routes having an
overlapping section, with the same end point forwarding node, among
routes between said arbitrary forwarding nodes.
[Claim 6] A control device, connected to a plurality of forwarding nodes for
which is set a processing rule that forwards a received packet in ac
cordance with a route set in advance, with a processing rule being
selected based on a path identifier included in said received packet, to
perform packet forwarding;
said control device comprising:
a first unit setting said processing rule beforehand in said forwarding
nodes in said route; and
a second unit causing a forwarding node positioned at a start point of
said route to add a path identifier in accordance with said route to a
received packet and causing a forwarding node positioned at an end
point of said route to delete said path identifier from a received packet,
on a predetermined occasion.
[Claim 7] The control device according to claim 6, wherein said control device
causes the forwarding nodes positioned at a start point and an end point
of said route to execute addition and deletion of the path identifier in
accordance with said route, at an occasion when a new flow detection
notification is received from said forwarding node.
[Claim 8] The control device according to claim 6 or 7, further comprising:
a route information storage unit that manages plurality of routes formed
by said forwarding nodes and path identifiers associated with said
routes respectively;
a route information management unit that selects an appropriate route
from said route information storage unit, when a new flow is generated;
and
an action computation unit that sets a processing rule in advance, in a
forwarding node in a route selected from said route information storage
unit, and also sets a processing rule respectively executing addition and
deletion of the path identifier in accordance with said route, in
forwarding nodes positioned at a start point and an end point of said
selected route.
[Claim 9] The control device according to any one of claims 6 to 8, wherein said
control device performs re-computation of a route between arbitrary
forwarding nodes and sets a processing rule corresponding to said
route, each time a change of topology of a network formed by said
WO 2012/073409 PCT/JP2011/005106
forwarding nodes is detected.
[Claim 10] A communication method, comprising:
a step in which, when a new flow is generated, a forwarding node p o
sitioned at a start point of a route that has been computed in advance
adds a path identifier to a received packet, based on an instruction from
a control device, and then forwards to a subsequent hop;
a step in which a forwarding node that is in a route computed in
advance selects a processing rule based on said path identifier included
in a received packet, from among processing rules set in advance by
said control device, and forwards the received packet; and
a step in which said forwarding node positioned at an end point of said
route restores said received packet to a state before said path identifier
was added, based on an instruction from said control device, and then
forwards to a subsequent hop.