Embodiments of the disclosure relate to the communications field, and in particular, to a method for packet tunneling through a software defined network (SDN), a method of intelligently controlling flow of a packet through an SDN network, and a system for tunneling used in an SDN.
The conventional Software defined network (SDN) technologies, e.g., the Openflow protocol used to interface an SDN controller and an SDN switch, can only provide limited support to selected tunneling technologies or protocols, such as GRE, MPLS, VLAN. Many commonly used tunneling protocols, such as IP-in-IP tunnel, VXLAN, NVGRE, and etc., are typically not supported by the conventional art. Moreover, based on the conventional technologies, an SDN system implementation typically is customized to support a specific tunneling technology, e.g., making decisions solely based on local logic depending on a packet. If an SDN system implementation is to be used to support more tunneling technologies, the forwarding plan tends to become undesirably complicated and usually involve update with each added tunneling technology.
An objective of embodiments of the disclosure is to provide a communication method for packet tunneling through an SDN, a computer implemented method of intelligently controlling flow of a packet through an SDN network, and a system for tunneling used in an SDN.
The technical solutions of the embodiments of the disclosure include the following content:
A communication method for packet tunneling through an SDN includes: sending programmable instructions to an SDN controller from a processor executing an application program that includes said programmable instructions; wherein said programmable instruction comprises primitive operations regarding processing a packet for tunneling in accordance with a tunneling protocol; configuring a flow table by said first SDN controller in accordance with said programmable instructions; and processing and distributing said packet in accordance with said flow table by said SDN switch.
Optionally, where said programmable instructions are transparent to said SDN switch.
A system for tunneling used in an SDN includes: an SDN controller, configured to receive instructions comprise operations of adding tunneling information to and removing tunneling information from a packet in accordance with a tunneling protocol, said instructions sent from a processor executing an application program; and configure a flow table associated with an SDN switch in accordance with said instructions; and an SDN switch, coupled to said SDN controller and configured to perform actions based on said flow table to distribute said packet in accordance with said tunneling protocol through said SDN network.
A computer implemented method of intelligently controlling flow of a packet through an SDN network, said method comprising configuring primitive operations regarding processing a packet for tunneling in accordance with a tunneling protocol, wherein said primitive operations comprise adding tunneling information to and/or removing tunneling information from a packet in accordance with said tunneling protocol, wherein said primitive operations are used by an SDN controller to configure a flow table, and wherein said flow table is used by an SDN switch to perform said primitive operations to distribute and/or receive a packet in accordance with said tunneling protocol.
Optionally, where said primitive operations are configured as respective data structures.
Optionally, where said SDN switch generic with respect to tunneling protocols.
An advantage of the embodiments of the disclosure is that the method may enable an SDN system to support multiple tunneling technologies without complicating the implementation of an SDN switch, and may enable the SDN system to support new tunneling technologies that have not been supported by the conventional art without the need for updating the SDN switch, which implementing packet tunneling through an SDN and reducing complexity of the SDN switch.
To illustrate the technical solutions in the embodiments of the disclosure more clearly, the accompanying drawings required for describing the embodiments are briefly described in the following. Apparently, the accompanying drawings in the following description merely show some embodiments of the disclosure, and persons of ordinary skill in the art may still derive other drawings from the accompanying drawings without creative efforts.
To make the objectives, technical solutions, and advantages of the embodiments of the disclosure clearer, the following clearly and completely describes the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are a part rather than all of the embodiments of the present invention. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
Therefore, it would be advantageous to provide a method to enable an SDN system to support multiple tunneling technologies without complicating the implementation of an SDN switch. It would also be advantageous to enable an SDN system to support new tunneling technologies that have not been supported by the conventional art without the need for updating the SDN switch.
Accordingly, the embodiments of the present disclosure employ an application program comprising instructions to configure an SDN switch flow table with primitive operations to implement tunneling. The primitive operations comprise adding tunneling information to, or encapsulating, a packet at an entry-point of a tunnel, and removing the tunneling information from, or decapsulating, a packet at an exit-point of a tunnel. The tunneling information may comprise header information as dictated by specifications of respective well known packet formats. The primitive operations may comprise pushing and popping the header information to configure a tunneling packet. Embodiments of the present disclosure may also push the tunneling information to metadata. The primitive operations may be implemented as data structures in an application program. An SDN controller, e.g., a centralized controller, can configure a flow table within an individual SDN switch based on the primitive operations as well as other instructions provided by the application program. The individual SDN switch can then take actions with respect to the packet based on the flow table. Therefore, the details of tunneling technology are placed in the application programs, while SDN switches are used to implement a set of primitive operations. This mechanism advantageously offers a generalized and extendable approach to support a wide variety of tunneling technologies, including non-standard tunneling technologies
As will be appreciate by those skilled in the art, the present disclosure is not limited to any specific type of tunneling technology. To name a few, the present disclosure can support MPLS, PBB, IP-in-IP (IPv4-in-IPv6, IPv6-in-IPv4), GRE, VXLAN, NVGRE, GPRS, PPPoE, CAPWAP, and so on.
Table 1 provides exemplary data structures defining actions for pushing and popping a MAC header on a packet in accordance with an embodiment of the present disclosure. The push_mac operation can be performed at a tunnel entry point to add the header to a packet, and the pop_mac operation can be performed at a tunnel node exit-point to remove the header from the packet. The parameters, e.g., src_mac, dst_mac, ethertype, are defined as dictated by the MAC protocol. In some embodiment, the MAC header is pushed/popped as the outmost header of a packet, e.g., as a default.
Table 2 provides exemplary data structures defining operations pushing and popping a VLAN header on a packet in accordance with an embodiment of the present disclosure. The pushing data structure comprises the position of the packet to push the header, the Ethernet type, the VLAN ID, and the priority. The popping data structure comprises a position to pop the header, such as the outmost header or the header following the MAC header.
Table 3 provides exemplary data structures defining operations for pushing and popping an IP header on a packet in accordance with an embodiment of the present disclosure. In this example, the data structures do not include all the header fields dictated by a IP packet format. As will be appreciated by those skilled in the art, other parameters for additional header fields can be added to the data structure in some embodiments. In some embodiments, an SDN switch may be capable of filling in the default values for header fields unspecified in the data structures while performing the action. Alternatively, the values of those unspecified header fields can be derived from information included an inner layer header. For example, if the application program is configured for a IPv4-in-IPv6 tunnel, some header fields of the outer IPv6 packet header can be derived from the inner IPv4 packet header, such as TTL.
Table 4 provides exemplary data structures defining operations for pushing and popping a transportation layer (TCP/UDP) header on a packet in accordance with an embodiment of the present disclosure. For example, the position specifies where to push/pop the TCP/UDP header to a packet. For example, position=1 means pushing/popping a TCP/UDP header at the beginning of a packet, or position=0 means pushing/popping a TCP/UDP header following a MAC header or the VLAN header if present. In this example, the data structures do not include all the header fields dictated by a TCP/UDP packet format. As will be appreciated by those skilled in the art, other parameters for additional header fields can be added to the data structure in some embodiments.
Requisite tunneling information to be added to a packet can also be added to a packet as metadata in accordance with embodiments of the present disclosure. This approach is particularly useful to support experimental, non-standard, or future new tunneling technologies. In some embodiments, the content of the metadata can be transparent to an SDN Switch. Thus a generic SDN switch can be used to support multiple tunneling protocols and new tunneling protocols, which may significantly simplify the implementation of the SDN switch, and circumvent the need for SDN switch update or upgrade in order to support new tunneling technologies. However, as will be appreciated by those skilled in the art, the present disclosure is not limited to any specific type of SDN switch. The SDN switch may only be generic with respect to selected tunneling protocols. It may include local logic customized for selected tunneling technologies but still capable of support other tunneling technologies in accordance with the primitive operations included in the API.
Table 5 provides exemplary data structures defining operations for pushing, popping and modifying metadata on a packet in accordance with an embodiment of the present disclosure. The parameter “mask” can be used to specify which bits to write/read. For example, if the total length is 64 bits and lower 16 bits are to be written, then the mask value can be 0x00 00 00 00 00 00 FF. The structure push_metadata comprises the position, length, and metadata content. In some embodiments, the position parameter can be selected from
0 default (behind mac, vlan, ip, tcp/udp)
1 Starting point of a packet
2 Behind MAC
3 Behind VLAN
4 Behind IP
5 Behind TCP/UDP
The structure metadata_to_flow comprises parameters for extracting values from metadata and use them for flow table match. The structure metadata_to_packet defines parameters for modifying metadata section in a packet, for example change the a GRE packet.
When handling a packet encapsulated inside a tunneling header with metadata in it, part or all of the metadata in the header can be put into metadata field in a flow structure for flow table lookup. Table 6 is an exemplary flow structure that includes metadata field in accordance with an embodiment of the present disclosure. Such a data structure, or part of it, can be used as a key for flow table lookup. The metadata field in the structure can have fixed or variable length.
When an SDN switch receives a packet, it extracts relevant information from the packet for flow table look up, for example the information in the MAC header or IP header. For a field that can not be recognized by the SDN switch, e.g., the metadata field, the SDN switch may not be able to process the field and the corresponding tunneling protocols. Thus, the tunneling information as part of the metadata can be added to the flow table lookup to process, for example based on the TNI in the NVGRE or the VNI in VxLAN, without adding all the metadata to the lookup table.
If an SDN switch receives a packet comprising a PBB encapsulation, it can perform a decapsulation process as shown in Table 8. As will be appreciated by those skilled in the art, any other suitable flow structure can be used to implement the similar method in accordance with the present disclosure.
IP-in-IP tunnels, including IPv4-in-IPv6, IPv6-in-IPv4 are widely used in IPv6 transition technologies. For example, DS-Lite, MAP-E, LW4over6, 4RD, etc, make use of IPv4-in-IPv6 tunnel; and 6RD makes use of IPv6-in-IPv4 tunnel.
Table 9 provides exemplary actions in an upstream direction and a downstream direction of a DS-Lite tunnel, respectively, in accordance with an embodiment of the present disclosure.
Table 10 provides exemplary instructions for push actions used in a IP-in-IP tunnel in accordance with an embodiment of the present disclosure. In this embodiment, when pushing an IP header, by default it is inserted behind MAC header; if VLAN header presents, it is inserted behind VLAN header. The Ethertype field in MAC header or VLAN header can be set to 0x800 if IPv4 header is inserted, and 0x86DD if IPv6 header is inserted.
If the payload above the inserted IP header is IPv4, then the protocol field of inserted IPv4 header or next-header field of IPv6 is set to 4. If the payload above the inserted IP header is IPv6, then the protocol field of inserted IPv4 header or next-header field of IPv6 can be set to 41. Other fields of inserted IP header can be derived from the IP header in payload, or take a default value. If the inserted IP header and the original IP header belong to different address families, e.g. IPv4 and IPv6, a mapping process can be used, which may result in loss of some field. The flow-label in IPv6 may not be mapped to IPv4 header.
Table 11 provides exemplary instructions for pop actions used in an IP-in-IP tunnel in accordance with an embodiment of the present disclosure. When an IP header is removed, the Ethertype field in the underlying MAC or VLAN header should be updated accordingly.
In the embodiments of the disclosure, GRE may be fully spelled as Generic Routing Encapsulation in English, and GRE may be the abbreviation of generic routing encapsulation in English. MPLS may be fully spelled as Multiprotocol Label Switching in English, and MPLS may be the abbreviation of Multiprotocol Label Switching in English. IP may be fully spelled as Internet Protocol in English, and IP may be the abbreviation of Internet Protocol in English. VXLAN may be fully spelled as Virtual Extensible Local Area Network in English, and VXLAN may be the abbreviation of Virtual Extensible Local Area Network in English. NVGRE may be fully spelled as Network Virtualization using Generic Routing Encapsulation in English, and NVGRE may be the abbreviation of Network Virtualization using Generic Routing Encapsulation in English. IPv4 may be fully spelled as Internet Protocol version 4 in English, and IPv4 may be the abbreviation of Internet Protocol version 4 in English. IPv6 may be fully spelled as Internet Protocol version 6 in English, and IPv6 may be the abbreviation of Internet Protocol version 6 in English. PBB may be fully spelled as Provider Backbone Bridge in English, and PBB may be the abbreviation of Provider Backbone Bridge in English.
In the embodiments of the disclosure, GPRS may be fully spelled as general packet radio service in English, and GPRS may be the abbreviation of general packet radio service in English. PPPoE may be fully spelled as Point-to-Point Protocol over Ethernet in English, and PPPoE may be the abbreviation of Point-to-Point Protocol over Ethernet in English. CAPWAP may be fully spelled as Control And Provisioning of Wireless Access Points in English, and CAPWAP may be the abbreviation of Control And Provisioning of Wireless Access Points in English. MAC may be fully spelled as Media Access Control in English, and MAC may be the abbreviation of Media Access Control in English. VLAN may be fully spelled as virtual local area network in English, and VLAN may be the abbreviation of virtual local area network in English. TTL may be fully spelled as time to live in English, and TTL may be the abbreviation of time to live in English. TCP may be fully spelled as Transmission Control Protocol in English, and TCP may be the abbreviation of Transmission Control Protocol in English. UDP may be fully spelled as User Datagram Protocol in English, and UDP may be the abbreviation of User Datagram Protocol in English.
In the embodiments of the disclosure, API may be fully spelled as application programming interface in English, and API may be the abbreviation of application programming interface in English. TNI may be fully spelled as Tenant Network Identifier in English, and TNI may be the abbreviation of Tenant Network Identifier in English. VNI may be fully spelled as VXLAN Network Identifier in English, and VNI may be the abbreviation of VXLAN Network Identifier in English. I-SID may be fully spelled as Backbone Service Instance Identifier in English, and I-SID may be the abbreviation of Backbone Service Instance Identifier in English. DS-Lite may be fully spelled as Dual Stack Lite in English, and DS-Lite may be the abbreviation of Dual Stack Lite in English. MAP-E may be fully spelled as Mapping of Address and Port with Encapsulation in English, and MAP-E may be the abbreviation of Mapping of Address and Port with Encapsulation in English. LW4over6 may be fully spelled as light weight 4 over 6 in English, and LW4over6 may be the abbreviation of light weight 4 over 6 in English. 4RD may be fully spelled as IPv4 Residual Deployment in English, and 4RD may be the abbreviation of IPv4 Residual Deployment in English. 6RD may be fully spelled as IPv6 Rapid Deployment in English, and 6RD may be the abbreviation of IPv6 Rapid Deployment in English.
A person of ordinary skill in the art may understand that all or a part of the steps of the method embodiments may be implemented by a program instructing relevant hardware. The program may be stored in a computer readable storage medium. When the program runs, the steps of the method embodiments are performed. The foregoing storage medium includes: any medium that can store program code, such as a ROM, a RAM, a magnetic disk, or an optical disc.
The foregoing descriptions are merely exemplary specific embodiments of the disclosure, but are not intended to limit the protection scope of the present invention. Any variation or replacement readily figured out by a person skilled in the art within the technical scope disclosed in the present invention shall fall within the protection scope of the present invention.
This application is a continuation of International Patent Application No. PCT/CN2014/083720, filed on Aug. 5, 2014, which claims priority to U.S. Provisional Application No. 61/862,400, filed on Aug. 5, 2013, both of which are hereby incorporated by reference in their entireties.
Number | Date | Country | |
---|---|---|---|
Parent | PCT/CN2014/083720 | Aug 2014 | US |
Child | 15006829 | US | |
Parent | 61862400 | Aug 2013 | US |
Child | PCT/CN2014/083720 | US |