Dynamic orchestration of overlay tunnels

Abstract

One embodiment of the present invention provides a dynamic overlay tunnel orchestration system. During operation, the system detects the appearance of a first virtual machine running on a hypervisor of a first host machine coupled to a first switch in a network. The system identifies a first virtual local area network (VLAN) associated with the first virtual machine and determines whether an overlay tunnel exists between the first switch and a second switch coupling a second virtual machine belonging to the first VLAN. If no such overlay tunnel exists, the system generates an instruction for the first and second switches to establish a first overlay tunnel between the first and second switches.

Description

BACKGROUND

Field

This disclosure relates to communication networks. More specifically, this disclosure relates to a system and method for dynamically establishing overlay tunnels.

Related Art

The exponential growth of the Internet has made it a popular delivery medium for a variety of applications running on physical and virtual devices. Such applications have brought with them an increasing demand for bandwidth. As a result, equipment vendors race to build larger and faster switches with versatile capabilities, such as network virtualization and multi-tenancy, to accommodate a plurality of tenant networks. Since Internet traffic is becoming more diverse, network virtualization is becoming progressively more important as a value proposition for network architects. In addition, the evolution of virtual computing has made multi-tenancy attractive and, consequently, placed additional requirements on the network. For example, virtual servers are being allocated to a large number of tenants while a respective tenant operates multiple virtualized networks. It is often desirable that the network infrastructure can provide a large number of virtualized networks to support multi-tenancy and ensure network separation among the tenants.

A flexible way of implementing network virtualization and multi-tenancy is using overlay networks for extending a tenant network. Network virtualization edge is a network entity that implements network virtualization functions. Typically, the network virtualization edge is implemented in host machines, which host virtualized servers (e.g., virtual machines). However, the number of virtualized networks cannot grow infinitely in a host machine. It is limited by processing capacity and design complexity, to name a few factors. Furthermore, host machines with higher capability are usually more complex and expensive. More importantly, host machines often cannot provide large number of virtualized networks due to the limited capability of a virtualization manager (e.g., a hypervisor running the virtual machines).

While network virtualization supports many desirable features, some issues remain unsolved in efficiently facilitating virtualized networks for multiple tenants.

SUMMARY

One embodiment of the present invention provides a dynamic overlay tunnel orchestration system. During operation, the system detects the appearance of a first virtual machine running on a hypervisor of a first host machine coupled to a first switch in a network. The system identifies a first virtual local area network (VLAN) associated with the first virtual machine and determines whether an overlay tunnel exists between the first switch and a second switch coupling a second virtual machine belonging to the first VLAN. If no such overlay tunnel exists, the system generates an instruction for the first and second switches to establish a first overlay tunnel between the first and second switches.

In a variation on this embodiment, the system allocates a network identifier for the first VLAN. The network identifier indicates traffic belonging to the first VLAN in the first overlay tunnel.

In a further variation, the system maintains a mapping between the first VLAN and a tuple comprising the first overlay tunnel and the network identifier.

In a variation on this embodiment, the system refrains from instructing the first and second switches to establish the first overlay tunnel in response to determining that an overlay tunnel exists between the first and second switches.

In a variation on this embodiment, the system detects the removal of the first virtual machine from the first host machine and determines whether the first overlay tunnel provides an overlay service to the first and second switches. If the first overlay tunnel does not provide an overlay service, the system generates an instruction for the first and second switches to terminate the first overlay tunnel between the first and second switches.

In a further variation, determining whether the first overlay tunnel provides an overlay service to the first and second switches includes one or more of: determining whether the first overlay tunnel carries traffic belonging to a second VLAN, and determining whether a third virtual machine belonging to the first VLAN is running on the hypervisor of the first host machine.

In a further variation, the removal of the first virtual machine includes: migration of the first virtual machine; and deletion of the first virtual machine.

In a variation on this embodiment, the system obtains information associated with the first virtual machine from a notification message from a notification agent in the hypervisor of the first host machine.

One embodiment of the present invention provides a switch. The switch includes a storage device, a tunnel management module, and a mapping module. During operation, the tunnel management module establishes an overlay tunnel with a second switch based on a first instruction identifying the tunnel endpoint switches of the overlay tunnel. The mapping module maintains, in the storage device, a mapping between a first tuple and a second tuple. The first tuple includes a media access control (MAC) address and a virtual local area network (VLAN) identifier of a virtual machine. The second tuple includes the overlay tunnel and a network identifier. The network identifier indicates traffic belonging to the VLAN in the overlay tunnel.

In a variation on this embodiment, the virtual machine is coupled to the second switch. The first instruction includes the VLAN identifier, the overlay tunnel, and the network identifier.

In a variation on this embodiment, the switch further includes a forwarding module, which encapsulates a packet in a tunnel encapsulation header associated with the overlay tunnel based on the mapping. The tunnel encapsulation header includes the network identifier.

In a variation on this embodiment, the tunnel management module terminates the overlay tunnel with the second switch based on a second instruction identifying tunnel endpoint switches for the overlay tunnel.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1A illustrates an exemplary network with support for dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention.

FIG. 1B illustrates an exemplary facilitation of dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention.

FIG. 2 illustrates exemplary mappings for facilitating dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention.

FIG. 3 illustrates an exemplary termination of a tunnel based on dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention.

FIG. 4A presents a flowchart illustrating the process of an overlay module establishing an overlay tunnel in response to detecting a new virtual machine, in accordance with an embodiment of the present invention.

FIG. 4B presents a flowchart illustrating the process of an overlay module terminating an overlay tunnel in response to removal of a virtual machine, in accordance with an embodiment of the present invention.

FIG. 5A presents a flowchart illustrating the process of a switch establishing an overlay tunnel, in accordance with an embodiment of the present invention.

FIG. 5B presents a flowchart illustrating the process of a switch terminating an overlay tunnel, in accordance with an embodiment of the present invention.

FIG. 6 presents a flowchart illustrating the process of a switch forwarding a packet via an overlay tunnel, in accordance with an embodiment of the present invention.

FIG. 7 illustrates an exemplary computer system and an exemplary switch for facilitating dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention.

In the figures, like reference numerals refer to the same figure elements.

DETAILED DESCRIPTION

The following description is presented to enable any person skilled in the art to make and use the invention, and is provided in the context of a particular application and its requirements. Various modifications to the disclosed embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the present invention. Thus, the present invention is not limited to the embodiments shown, but is to be accorded the widest scope consistent with the claims.

Overview

In embodiments of the present invention, the problem of efficiently facilitating overlay tunnels for virtual machines in a network is solved by dynamically orchestrating an overlay tunnel between switch pairs coupling virtual machines belonging to the same layer-2 broadcast domain (e.g., the same virtual local area network (VLAN)). Examples of a tunnel include, but are not limited to, Virtual Extensible Local Area Network (VXLAN); Generic Routing Encapsulation (GRE); and GRE variants, such as Network Virtualization using GRE (NVGRE) and openvSwitch GRE.

Suppose that two virtual machines residing on two different host machines belong to a same VLAN but are separated by a network (e.g., a layer-3 network). With existing technologies, the virtualization managers (e.g., hypervisors and Hyper-Vs) running the virtual machines establish an overlay tunnel to extend the VLAN across the network. However, the capacity of a virtualization manager is limited by the processing capacity of the host machine hosting the virtualization manager and the virtual machines. To overcome such limitations, the responsibilities of managing and maintaining overlay tunnels can be offloaded to the switches of the network. In traditional overlay networks, the orchestration of overlay tunnels among the switches are static and pre-determined based on the network architecture. The switches are configured to establish the overlay tunnels between a respective switch pair in the network during the boot up process. These tunnels typically do not change, even when new virtual machines are added or existing ones are removed or migrated.

To solve this problem, since the overlay tunnels are managed by the switches, the responsibilities of encapsulating and decapsulating packets with tunnel headers is offloaded from the virtualization managers to the switches. In some embodiments, the switches use switch modules (e.g., switch hardware) to manage the tunnels. As a result, the switches can encapsulate and decapsulate packets more efficiently than the virtualization managers.

In addition, an overlay module running on a configuration manager can collect information about appearance and removal of virtual machines on the virtualization managers. For example, a virtualization manager can include a notification agent, which provides the information to the overlay module. Based on this information, the overlay module dynamically determines whether to establish or terminate a tunnel between a switch pair. The overlay module determines the necessity of an overlay tunnel based on the lifecycle of the virtual machines. The overlay module establishes an overlay tunnel between any two switches when both switches are coupled to virtual machines requiring overlay services (e.g., virtual machines belonging to a same VLAN). The overlay module removes a tunnel when a virtual machine no longer requires the services provided by the overlay tunnel (e.g., when a virtual machine is removed).

In some embodiments, the network is a fabric switch. In a fabric switch, any number of switches coupled in an arbitrary topology may logically operate as a single switch. The fabric switch can be an Ethernet fabric switch or a virtual cluster switch (VCS), which can operate as a single Ethernet switch. Any member switch may join or leave the fabric switch in “plug-and-play” mode without any manual configuration. In some embodiments, a respective switch in the fabric switch is a Transparent Interconnection of Lots of Links (TRILL) routing bridge (RBridge). In some further embodiments, a respective switch in the fabric switch is an IP routing-capable switch (e.g., an IP router).

It should be noted that a fabric switch is not the same as conventional switch stacking. In switch stacking, multiple switches are interconnected at a common location (often within the same rack), based on a particular topology, and manually configured in a particular way. These stacked switches typically share a common address, e.g., an IP address, so they can be addressed as a single switch externally. Furthermore, switch stacking requires a significant amount of manual configuration of the ports and inter-switch links. The need for manual configuration prohibits switch stacking from being a viable option in building a large-scale switching system. The topology restriction imposed by switch stacking also limits the number of switches that can be stacked. This is because it is very difficult, if not impossible, to design a stack topology that allows the overall switch bandwidth to scale adequately with the number of switch units.

In contrast, a fabric switch can include an arbitrary number of switches with individual addresses, can be based on an arbitrary physical topology, and does not require extensive manual configuration. The switches can reside in the same location, or be distributed over different locations. These features overcome the inherent limitations of switch stacking and make it possible to build a large “switch farm,” which can be treated as a single, logical switch. Due to the automatic configuration capabilities of the fabric switch, an individual physical switch can dynamically join or leave the fabric switch without disrupting services to the rest of the network.

Furthermore, the automatic and dynamic configurability of the fabric switch allows a network operator to build its switching system in a distributed and “pay-as-you-grow” fashion without sacrificing scalability. The fabric switch's ability to respond to changing network conditions makes it an ideal solution in a virtual computing environment, where network loads often change with time.

It should also be noted that a fabric switch is distinct from a VLAN. A fabric switch can accommodate a plurality of VLANs. A VLAN is typically identified by a VLAN tag. In contrast, the fabric switch is identified by a fabric identifier (e.g., a cluster identifier), which is assigned to the fabric switch. Since a fabric switch can be represented as a logical chassis, the fabric identifier can also be referred to as a logical chassis identifier. A respective member switch of the fabric switch is associated with the fabric identifier. In some embodiments, a fabric switch identifier is pre-assigned to a member switch. As a result, when the switch joins a fabric switch, other member switches identifies the switch to be a member switch of the fabric switch.

In this disclosure, the term “fabric switch” refers to a number of interconnected physical switches which can form a single, scalable network of switches. The member switches of the fabric switch can operate as individual switches. The member switches of the fabric switch can also operate as a single logical switch in the provision and control plane, the data plane, or both. “Fabric switch” should not be interpreted as limiting embodiments of the present invention to a plurality of switches operating as a single, logical switch. In this disclosure, the terms “fabric switch” and “fabric” are used interchangeably.

Although the present disclosure is presented using examples based on an encapsulation protocol, embodiments of the present invention are not limited to networks defined using one particular encapsulation protocol associated with a particular Open System Interconnection Reference Model (OSI reference model) layer. For example, embodiments of the present invention can also be applied to a multi-protocol label switching (MPLS) network. In this disclosure, the term “encapsulation” is used in a generic sense, and can refer to encapsulation in any networking layer, sub-layer, or a combination of networking layers.

The term “end device” can refer to any device external to a network (e.g., does not perform forwarding in that network). Examples of an end device include, but are not limited to, a physical or virtual machine, a conventional layer-2 switch, a layer-3 router, or any other type of network device. Additionally, an end device can be coupled to other switches or hosts further away from a layer-2 or layer-3 network. An end device can also be an aggregation point for a number of network devices to enter the network. An end device hosting one or more virtual machines can be referred to as a host machine. In this disclosure, the terms “end device” and “host machine” are used interchangeably.

The term “hypervisor” is used in a generic sense, and can refer to any virtualization manager (e.g., a virtual machine manager). Any software, firmware, or hardware that creates and runs virtual machines can be a “hypervisor.” The term “virtual machine” is also used in a generic sense and can refer to software implementation of a machine or device. Any virtual device which can execute a software program similar to a physical device can be a “virtual machine.”

The term “VLAN” is used in a generic sense, and can refer to any virtualized network. Any virtualized network comprising a segment of physical networking devices, software network resources, and network functionality can be can be referred to as a “VLAN.” “VLAN” should not be interpreted as limiting embodiments of the present invention to layer-2 networks. “VLAN” can be replaced by other terminologies referring to a virtualized network or network segment, such as “Virtual Private Network (VPN),” “Virtual Private LAN Service (VPLS),” or “Easy Virtual Network (EVN).”

The term “packet” refers to a group of bits that can be transported together across a network. “Packet” should not be interpreted as limiting embodiments of the present invention to layer-3 networks. “Packet” can be replaced by other terminologies referring to a group of bits, such as “frame,” “cell,” or “datagram.”

The term “switch” is used in a generic sense, and can refer to any standalone or fabric switch operating in any network layer. “Switch” can be a physical device or software running on a computing device. “Switch” should not be interpreted as limiting embodiments of the present invention to layer-2 networks. Any device that can forward traffic to an external device or another switch can be referred to as a “switch.” Examples of a “switch” include, but are not limited to, a layer-2 switch, a layer-3 router, a TRILL RBridge, or a fabric switch comprising a plurality of similar or heterogeneous smaller physical switches.

The term “edge port” refers to a port on a network which exchanges data frames with a device outside of the network (i.e., an edge port is not used for exchanging data frames with another member switch of a network). The term “inter-switch port” refers to a port which sends/receives data frames among member switches of the network. A link between inter-switch ports is referred to as an “inter-switch link.” The terms “interface” and “port” are used interchangeably.

The term “switch identifier” refers to a group of bits that can be used to identify a switch. Examples of a switch identifier include, but are not limited to, a media access control (MAC) address, an Internet Protocol (IP) address, an RBridge identifier, or a combination thereof. In this disclosure, “switch identifier” is used as a generic term, is not limited to any bit format, and can refer to any format that can identify a switch.

The term “tunnel” refers to a data communication where one or more networking protocols are encapsulated using another networking protocol. Although the present disclosure is presented using examples based on a layer-3 encapsulation of a layer-2 protocol, “tunnel” should not be interpreted as limiting embodiments of the present invention to layer-2 and layer-3 protocols. A “tunnel” can be established for and using any networking layer, sub-layer, or a combination of networking layers.

Network Architecture

FIG. 1A illustrates an exemplary network with support for dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention. As illustrated in FIG. 1A, a network 100 includes member switches 101, 102, 103, 104, and 105. Network 100 can be a TRILL network and a respective member switch, such as switch 105, can be a TRILL RBridge. Network 100 can also be an IP network and a respective member switch, such as switch 105, can be an IP-capable switch, which calculates and maintains a local IP routing table (e.g., a routing information base or RIB), and is capable of forwarding packets based on its IP addresses. In some embodiments, network 100 is a fabric switch, and one or more switches in fabric switch 100 can be virtual switches (e.g., a software switch running on a computing device).

Switch 103 is coupled to end devices 120 and 130. Switches 102 and 105 are coupled to end devices 140 and 150, respectively. End devices 120, 130, 140, and 150 are host machines, each hosting a plurality of virtual machines. Switches in network 100 use edge ports to communicate with end devices and inter-switch ports to communicate with other switches. For example, switch 103 is coupled to end devices, such as end device 120, via edge ports and to switches 101, 102, and 104 via inter-switch ports. End devices 120, 130, 140, and 150 include hypervisors 122, 132, 142, and 152, respectively. Virtual machines (VMs) 124 and 126 run on hypervisor 122; virtual machines 134 and 136 run on hypervisor 132; virtual machines 144 and 146 run on hypervisor 142; and virtual machines 154 and 156 run on hypervisor 152.

In this example, virtual machines 124, 134, 144, 146, and 154 belong to a tenant 1 and virtual machines 126, 136, and 156 belong to a tenant 2. Tenant 1 deploys VLANs 112 and 114, and tenant 2 deploys VLAN 112. Hence, the same VLAN identifier for VLAN 112 can be used by multiple tenants. Virtual machines 124 and 144 are in VLAN 112 of tenant 1, virtual machines 134, 146, and 154 are in VLAN 114 of tenant 1, and virtual machines 126, 136, and 156 are in VLAN 112 of tenant 2. Since network 100 is serving a plurality of tenants, each deploying a plurality of VLANs, a respective switch of network 100 can serve both tenants 1 and 2, and a plurality of switches can serve the same tenant 1 or 2.

With existing technologies, since virtual machines 124 and 144 belong to a same VLAN 112 but are separated by network 100, hypervisors 122 and 142 establish an overlay tunnel to extend VLAN 112 across network 100. However, the capacity of hypervisors 122 and 142 is limited by the processing capacity of host machines 120 and 140, respectively. To overcome such limitations, the responsibilities of managing and maintaining overlay tunnels can be offloaded to the switches. In traditional overlay networks, the switches of network 100 are configured to establish the overlay tunnels between a respective switch pair in network 100 during the boot up process. These tunnels typically do not change, even when new virtual machines are added or existing ones are removed or migrated.

To solve this problem, an end device 110 operating as a configuration manager can dynamically orchestrate overlay tunnels in network 100 between switches coupling virtual machines belonging to the same VLAN. These switches can be referred to as tunnel endpoints. A tunnel is identified by the switch identifiers (e.g., IP addresses) of the tunnel endpoints. Configuration manager 110 can be an administrator device from which network 100 can be configured. End device 110 can be coupled with network 100 via one or more links (denoted with a dashed line) of one or more networks. Since the responsibility of managing overlay tunnels is offloaded to the switches of network 100, the responsibilities of encapsulating and decapsulating packets with tunnel headers is offloaded from hypervisors to the switches in network 100. In some embodiments, the switches use switch hardware to manage the tunnels. As a result, the switches can encapsulate and decapsulate packets more efficiently than the hypervisors.

In some embodiments, an overlay module 172 running on configuration manager 110 can collect information about appearance and removal of virtual machines. For example, when virtual machines 124 and 144 are created (or instantiated) on hypervisors 122 and 142, overlay module 172 collects this information. Based on this information, overlay module 172 determines that virtual machines 124 and 144 belong to a same VLAN 112 of tenant 1. As a result, overlay module 172 dynamically determines to establish an overlay tunnel between switches 102 and 103. Overlay module 172 sends a control message to switches 102 and 104 instructing them to establish an overlay tunnel. Upon receiving the control message, switches 102 and 103 establish a tunnel 162 (e.g., a VXLAN tunnel) between them.

In the same way, overlay module 172 determines that virtual machines 134 and 154 belong to a same VLAN 114 of tenant 1. Hence, based on an instruction from overlay module 172, switches 103 and 105 establish a tunnel 164 between them. This instruction includes the identifying information of switches 103 and 105. Furthermore, since switch 102 is coupled to a virtual machine 146 belonging to VLAN 114 of tenant 1, switches 102 and 103 use an already established tunnel 162 to carry traffic of VLAN 114. In the same way, overlay module 172 determines that virtual machines 126, 136, and 156 belong to a same VLAN 112 of tenant 2. Switches 103 and 105 then use already established tunnel 164 to carry traffic of VLAN 112 of tenant 2. Hence, the same tunnel can be used to carry traffic of different tenants. In this example, since switches 102 and 105 don't share a layer-2 broadcast domain, overlay module 172 determines that switches 102 and 105 may not need an overlay tunnel to forward traffic between host machines 140 and 150.

It should be noted that even though the same tunnel 162 can carry traffic of both VLANs 112 and 114 of tenant 1, these VLANs can use separate network identifiers (e.g., VXLAN Network Identifiers or VNIs). In other words, packets encapsulated with tunnel encapsulation headers carry different network identifiers for packets of VLANs 112 and 114. In the same way, even though the same tunnel 164 can carry traffic of both VLAN 114 of tenant 1 and VLAN 112 of tenant 2, these VLANs use separate network identifiers. The use of separate network identifiers allows the switches to use the same tunnel and maintain VLAN separations.

Overlay module 172 determines the necessity of overlay tunnel 162 based on the lifecycle of virtual machines coupled to switches 102 and 103. Suppose that virtual machines 144 and 146 are not coupled with switch 102 (e.g., due to a migration or a removal). Overlay module 172 determines that switch 102 does not have any virtual machine coupled with it. As a result, overlay module 172 terminates tunnel 162 since virtual machines 144 and 146 no longer require the services provided by overlay tunnel 162, such as extending VLANs 112 and 114 of tenant 1 across network 100.

FIG. 1B illustrates an exemplary facilitation of dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention. In this example, configuration manager 110 runs a virtual machine management module 170. In some embodiments, virtual machine management module 170 is a cloud networking controller (e.g., OpenStack Neutron). Virtual machine management module 170 can provide flow definitions to the switches of network 100 (e.g., flow definitions in a software-defined network). The switches in network 100 can process packets based on the flow definition. Virtual machine management module 170 can include one or more application program interfaces (APIs), plug-ins, and authentication/authorization control modules that facilitates orchestration of virtual machines. Overlay module 172 can be a plug-in running on virtual machine management module 170.

To facilitate information to virtual machine management module 170, hypervisors 132 and 152 run notification agents 182 and 184, respectively. In some embodiments, notification agents 182 and 184 facilitate a messaging architecture (e.g., OpenStack Nova) for virtual machine management module 170 and can send messages to virtual machine management module 170 regarding appearance and removal of virtual machines on hypervisors 132 and 152, respectively. When virtual machine management module 170 receives such messages, overlay module 172 obtains the information provided in the messages. It should be noted that overlay module 172 only obtains information regarding the virtual machines from virtual machine management module 170. The overlay tunnel orchestration facilitated by overlay module 172 can be a capability independent of the native capabilities of virtual machine management module 170.

In addition, when a virtual machine is added to or removed from host machine 130, hypervisor 132 sends a discovery message 180 to switch 103. In this way, switch 103 identifies the media access control (MAC) address of the virtual machine. In some embodiments, discovery message 180 is based on a Link Layer Discovery Protocol (LLDP). Switch 103 can further construct a notification message indicating that switch 103 has learned the MAC addresses of virtual machines 134 and 136 and sends the notification message to switch 105. In this way, switch 105 determines that virtual machines 134 and 136 are reachable via switch 103. In the same way, switch 105 discovers appearance or removal of a virtual machine in host machine 150 and shares the information with switch 103.

Based on the information provided by notification agents 182 and 184, overlay module 172 dynamically determines whether to establish or terminate a tunnel between a switch pair. During operation, notification agent 182 detects that a virtual machine 124 has appeared in host machine 120. Notification agent 182 then sends a message comprising the virtual MAC address and VLAN 114 of virtual machine 124 to virtual machine management module 170. When the message reaches virtual machine management module 170, overlay module 172 obtains the information. Overlay module 172 determines that there is no other virtual machine in VLAN 114 of tenant 1 coupled with network 100. However, when virtual machine 154 appears in host machine 150, overlay module 172 obtains the MAC address and VLAN 114 of virtual machine 154. Overlay module 172 determines that both virtual machines 124 and 154 belong to the same VLAN 114 of tenant 1.

Based on this information, overlay module 172 determines that host machine 120 is coupled to switch 103 and host machine 150 is coupled to switch 105. Overlay module 172 generates a notification message instructing switches 103 and 105 to establish an overlay tunnel. In some embodiments, an overlay tunnel formation is disabled for hypervisors 122 and 152 (e.g., by an administrator). This allows a hypervisor to forward a layer-2 frame to a switch. Overlay module 172 can assign a network identifier for VLAN 114 of tenant 1, maintain a mapping between them, and send the mapping to switches 103 and 105. Upon receiving the instruction, switches 103 and 105 establish tunnel 164 between them.

On the other hand, when virtual machines 126 and 156 are created, notification agents 182 and 184, respectively, notify virtual machine management module 170. Overlay module 172 obtains the information regarding virtual machines 126 and 156, and determines that virtual machines 126 and 156 belong to VLAN 112 of tenant 2. Overlay module 172 determines that since tunnel 164 has already been established between switches 103 and 105, another tunnel is not needed to forward packets between hypervisors 122 and 124.

To forward a packet to virtual machine 154, virtual machine 124 provides the packet to hypervisor 122, which in turn, forwards the packet to switch 103. Switch 103 encapsulates the packet in a tunnel encapsulation header, includes the network identifier in the encapsulation header, and forwards the encapsulated packet to switch 105 via tunnel 164. The network identifier allows separation of traffic in network 100 between VLAN 114 of tenant 1 and VLAN 112 of tenant 2.

Tunnel Mapping

FIG. 2 illustrates exemplary mappings for facilitating dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention. During operation, overlay module 172 discovers information associated with a respective virtual machine coupled to network 100. Discovery information for a virtual machine includes a MAC address and a VLAN. Overlay module 172 allocates a network identifier (e.g., a VNI) for a respective VLAN so that a respective VLAN can be uniquely identified in an overlay tunnel. Overlay module 172 maintains a tunnel mapping comprising the tunnel, the VNI, and the VLAN information. The VLAN information includes a VLAN, which is identified by a VLAN identifier (e.g., an Institute of Electrical and Electronics Engineers (IEEE) 802.1Q VLAN tag or its variation), and a tenant of the VLAN. A tenant can be identified in the mapping based on one or more of: a tenant identifier, an ingress port, an egress port, and an IP subnet.

For example, since tunnel 162 carries traffic belonging to VLAN 112 of tenant 1, overlay module 172 maintains a tunnel mapping 202, which maps VLAN 112 and tenant 1 to tunnel 162 and a VNI 222 allocated for VLAN 112 of tenant 1. In tunnel mapping 202, tunnel 162 is identified by switch identifiers of the tunnel endpoints of tunnel 162. In other words, tunnel mapping 202 maps a tuple comprising VLAN 112 and tenant 1 to another tuple comprising tunnel 162 and VNI 222. Similarly, since tunnel 162 also carries traffic belonging to VLAN 114 of tenant 1, tunnel mapping 202 further maps VLAN 114 and tenant 1 to tunnel 162 and a VNI 224 allocated for VLAN 114 of tenant 1. VNIs 222 and 224 distinguish traffic belonging to VLANs 112 and 114, respectively, in tunnel 162. Tunnel mapping 202 allows the same tunnel 162 to carry traffic of both VLANs 112 and 114 of tenant 1 by using VNIs 222 and 224, respectively, in the tunnel encapsulation header (e.g., a VXLAN header).

In the same way, since tunnel 164 carries traffic belonging to VLAN 114 of tenant 1, tunnel mapping 202 further maps VLAN 114 and tenant 1 to tunnel 164 and a VNI 226 allocated for VLAN 114 of tenant 1. Similarly, since tunnel 164 also carries traffic belonging to VLAN 112 of tenant 2, tunnel mapping 202 further maps VLAN 112 and tenant 2 to tunnel 164 and a VNI 228 allocated for VLAN 112 of tenant 2. VNIs 226 and 228 distinguish traffic belonging to VLAN 114 of tenant 1 and VLAN 112 of tenant 2, respectively, in tunnel 164. It should be noted that overlay module 172 allocates both VNIs 224 and 226 for VLAN 114 of tenant 1, but for tunnels 162 and 164, respectively. In some embodiments, VNIs 224 and 226 have the same identifier value.

Overlay module 172 constructs a notification message comprising tunnel mapping 202 and sends the notification message to a switch, such as switch 103, in network 100. In some embodiments, overlay module 172 only includes the mapping of the tunnels for which switch 103 is a tunnel endpoint (i.e., the tunnels initiated or terminated at switch 103) in the notification message. When switch 103 learns a MAC address of a remote virtual machine, which is not coupled to switch 103, switch 103 creates a forwarding mapping 206 for that MAC address.

Suppose that MAC addresses 252, 254, 256, and 258 belong to virtual machines 144, 146, 154, and 156, respectively. When switch 103 learns MAC address 252 (e.g., either from a local port or from a notification message from switch 102), switch 103 determines that virtual machine 144 belongs to VLAN 112 of tenant 1. Based on the mappings received from overlay module 172, switch 103 creates an entry in forwarding mapping 206 for MAC address 252. Since tunnel 162 carries traffic from switch 103 to virtual machine 144, forwarding mapping 206 maps MAC address 252, VLAN 112, and tenant 1 to tunnel 162 and VNI 222 based on the mappings received from overlay module 172. In other words, tunnel mapping 206 maps a tuple comprising MAC address 252, VLAN 112, and tenant 1 to another tuple comprising tunnel 162 and VNI 222.

In the same way, forwarding mapping 206 maps MAC address 254, VLAN 114, and tenant 1 to tunnel 162 and VNI 224; MAC address 256, VLAN 114, and tenant 1 to tunnel 164 and VNI 226; and MAC address 258, VLAN 112, and tenant 2 to tunnel 164 and VNI 228. Forwarding mapping 206 allows switch 103 to determine which tunnel and VNI to use for forwarding packets via the overlay tunnels. Since a tunnel is identified based on the switch identifiers (e.g., IP addresses) of the tunnel endpoints, switch 103 can determine the destination address for a tunnel encapsulation header from forwarding mapping 206.

For example, upon receiving a packet destined to virtual machine 156 from virtual machine 136, switch 103 obtains forwarding information from forwarding mapping 206. Switch 103 encapsulates the packet with a tunnel encapsulation header, sets the source and destination addresses of the encapsulation header as the switch identifiers (e.g., IP addresses) of switches 103 and 105, respectively, and includes VNI 228 as the network identifier of the encapsulation header. Switch 103 can obtain an egress port for the encapsulated packed based on the destination address of the encapsulation header and transmit the packet via the egress port.

Tunnel Termination

Overlay module 172 determines the necessity of an overlay tunnel based on the lifecycle of the virtual machines. Hence, overlay module 172 removes a tunnel when a virtual machine no longer requires the services provided by the overlay tunnel (e.g., when a virtual machine is removed). FIG. 3 illustrates an exemplary termination of a tunnel based on dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention. During operation, virtual machine 156 migrates to host machine 140 (denoted with a dotted arrow) and starts running on hypervisor 142. Virtual machine 156 then becomes coupled to switch 102 but still remains in VLAN 112 of tenant 2.

Overlay module 172 obtains this information from notification agent 184. Overlay module 172 determines virtual machine 156 to be a new virtual machine running on hypervisor 142 and determines whether to establish a tunnel to accommodate the new virtual machine. Since virtual machines 126 and 136 are coupled to switch 103 and in VLAN 112 of tenant 2, tunnel 162 can carry traffic between migrated virtual machine 156 and virtual machine 126 (or 136). To facilitate this, in some embodiments, overlay module 172 updates tunnel mapping 202 to map VLAN 112 and tenant 2 to tunnel 162 and VNI 228, thereby associating VNI 228 with tunnel 162. Overlay module 172 can provide this updated information to switch 102.

Overlay module 172 then checks whether tunnel 164 is needed to forward traffic. Overlay module 172 determines that host machine 150 is still hosting virtual machine 154. As a result, the packets between virtual machines 134 and 154, which belong to the same VLAN 112 of tenant 2, are still forwarded via tunnel 164. Hence, overlay module 172 determines that tunnel 164 is needed to forward traffic. Switches 103 and 105 thus continue to maintain tunnel 164 between them. In this way, overlay module 172 ensures that even when a virtual machine migrates, the associated overlay tunnel continues carry traffic if another virtual machine needs the overlay tunnel.

Suppose that virtual machine 154 is removed (denoted with a cross). Overlay module 172 obtains this information from notification agent 184 and determines that host machine 150 is not hosting any virtual machine that needs an overlay tunnel. As a result, overlay module 172 determines that tunnel 164 is not needed to forward traffic. Overlay module 172 then generates a notification message instructing to terminate tunnel 164 and sends the notification message to switches 103 and 105. Upon receiving the notification message, switches 103 and 105 terminate tunnel 164, thereby removing an unused overlay tunnel from network 100.

Operations of an Overlay Module

FIG. 4A presents a flowchart illustrating the process of an overlay module establishing an overlay tunnel in response to detecting a new virtual machine, in accordance with an embodiment of the present invention. During operation, the overlay module detects the appearance of a new virtual machine (operation 402). The appearance of a new virtual machine can be triggered by migration of a virtual machine from another host machine and creation of a new virtual machine. The overlay module identifies a VLAN (e.g., a layer-2 broadcast domain) associated with the new virtual machine (operation 404). The overlay module then identifies the switch coupled to the new virtual machine (operation 406) and checks whether the same VLAN is configured in another switch (operation 408). If the same VLAN is configured in another switch, the overlay module checks whether a tunnel between the switch and the other switch already exists (operation 410).

If a tunnel between the switch and the other switch doesn't exist, the overlay module allocates a VNI for the VLAN for the tunnel between the switch and the other switch (operation 414). The overlay module generates a mapping between the VLAN and the (tunnel, VNI) tuple (operation 416). The mapping can further include a tenant to which the VLAN belong. The overlay module generates a control message comprising the mapping and an instruction for establishing the tunnel based on the mapping (operation 418). The overlay module then identifies respective egress ports for the switch and the other switch, and transmits the control message via the identified ports (operation 420).

If the overlay module has transmitted the control message (operation 420), the same VLAN is not configured in another switch (operation 408), or a tunnel between the switch and the other switch already exists (operation 410), the overlay module may not create an overlay tunnel for the other switch. It should be noted that the overlay module can allocate a VNI for a VLAN even when the tunnel exists. The overlay module then checks whether all switches have been checked (operation 412). If all switches have not been checked, the overlay module continues to check whether the same VLAN is configured in another switch (operation 408) and whether a tunnel between the switch and the other switch already exists (operation 410).

FIG. 4B presents a flowchart illustrating the process of an overlay module terminating an overlay tunnel in response to removal of a virtual machine, in accordance with an embodiment of the present invention. During operation, the overlay module detects the removal of a virtual machine (operation 452). The removal includes migration and deletion of the virtual machine. The overlay module identifies a VLAN associated with the virtual machine (operation 454). The overlay module then identifies the switch coupled to the new virtual machine (operation 456) and checks whether another virtual machine in the same VLAN is coupled to the switch (operation 458). If another virtual machine in the same VLAN is not coupled to the switch, the overlay module identifies the tunnel which has been carrying traffic of the VLAN (operation 460).

The overlay module checks whether the tunnel carries traffic of another VLAN (operation 462). In this way, the overlay module determines whether the tunnel still provides an overlay service to the tunnel endpoint switches. If the tunnel does not carry traffic of another VLAN, the tunnel can be terminated. The overlay module then removes the mapping between the VLAN and the (tunnel, VNI) tuple (operation 466). The overlay module generates a control message comprising the mapping and an instruction for terminating the tunnel based on the mapping (operation 468). The overlay module then identifies respective egress ports for tunnel endpoint switches and transmits the control message via the identified ports (operation 470).

If another virtual machine in the same VLAN is coupled to the switch (operation 458), the overlay module may not terminate the overlay tunnel. If the overlay module has transmitted the control message (operation 470) or another virtual machine in the same VLAN is coupled to the switch (operation 462), the overlay module checks whether all tunnels carrying the traffic of the VLAN have been checked (operation 464). If all tunnels carrying the traffic of the VLAN have not been checked, the overlay module continues to identify the next tunnel which has been carrying traffic of the VLAN (operation 460).

Operations of a Switch

FIG. 5A presents a flowchart illustrating the process of a switch establishing an overlay tunnel, in accordance with an embodiment of the present invention. During operation, the switch obtains a mapping and an instruction for establishing a tunnel from a control message (operation 502) and identifies the remote switch of the tunnel (e.g., the IP address of the remote switch) from the message (operation 504). The switch establishes the tunnel with the remote switch and associates the VNI from the mapping with the tunnel (operation 506). The switch identifies the VLAN from the mapping (operation 508) and obtains the MAC addresses, which have been learned from the remote switch, associated with the VLAN (operation 510). The switch then creates a local forwarding mapping between the learned MAC addresses and the (tunnel, VNI) tuple (operation 512).

FIG. 5B presents a flowchart illustrating the process of a switch terminating an overlay tunnel, in accordance with an embodiment of the present invention. During operation, the switch obtains a mapping and an instruction for terminating a tunnel from a control message (operation 552) and identifies the remote switch of the tunnel from the message (operation 554). The switch identifies the local forwarding mappings associated with the tunnel (operation 556) and removes the identified forwarding mappings (operation 558). The switch then terminates the tunnel with the remote switch (operation 560).

FIG. 6 presents a flowchart illustrating the process of a switch forwarding a packet via an overlay tunnel, in accordance with an embodiment of the present invention. During operation, the switch identifies the destination identifier of the packet received via a local edge port (operation 602) and identifies the (tunnel, VNI) tuple mapped to the destination MAC address of the packet (operation 604). The switch then encapsulates the packet with a tunnel encapsulation header and sets the address of the remote tunnel endpoint switch as the destination address of the tunnel encapsulation header (operation 606). The switch sets the VNI as the tunnel identifier in the tunnel encapsulation header (operation 608). The switch identifies the egress port associated with the remote tunnel endpoint switch (operation 610) and transmits encapsulated packet via the identified port (operation 612).

Exemplary Computer System and Switch

FIG. 7 illustrates an exemplary computer system and an exemplary switch for facilitating dynamic orchestration of overlay tunnels, in accordance with an embodiment of the present invention. In this example, a computer system 750 includes a number of communication ports 752, a processor 762, a memory 764, a packet processor 760, and a storage device 790. Memory 764 can include a volatile memory (e.g., RAM) that serves as a managed memory, and can be used to store one or more memory pools. Memory 764 stores instructions that when executed by processor 762 cause computer system 750 to perform methods and/or processes for facilitating dynamic orchestration of overlay tunnels in network 758, as described in this disclosure. Furthermore, computer system 750 can be coupled to a display device 754 and a keyboard 756.

Computer system 750 further includes a virtual machine management module 770, an overlay module 772, and a notification module 780. During operation, virtual machine management module 770 detects the appearance of a virtual machine running on a hypervisor of a host machine coupled to a switch 700. Virtual machine management module 770 can obtain such information from a notification agent in the hypervisor, as described in conjunction with FIG. 1B. Overlay module 772 identifies a VLAN associated with the virtual machine and determines whether an overlay tunnel exists between switch 700 and a remote switch coupling a another virtual machine belonging to the VLAN. If no such overlay tunnel exists, notification module 780, in conjunction with overlay module 772, generates an instruction for switch 700 and the remote switch to establish an overlay tunnel between the switches. Packet processor 760 forwards the instruction, which can be included in a message, to the switches. On the other hand, if an overlay tunnel exists, overlay module 772 refrains from instructing the switches to establish the overlay tunnel.

Overlay module 772 allocates a network identifier for the VLAN and maintains, in storage device 790, a mapping between the VLAN and a tuple comprising the overlay tunnel (e.g., identified by switch identifiers) and the network identifier. In some embodiments, overlay module 772 detects the removal of the virtual machine from the host machine and determines whether the overlay tunnel provides an overlay service to the switches. If not, overlay module 772 generates an instruction for the switches to terminate the overlay tunnel.

In some embodiments, switch 700 includes a fabric switch module 724, which maintains a membership in a network of interconnected switches. A respective switch of the network is associated with a group identifier identifying the network. Switch 700 then maintains a configuration database in a storage device 740 that maintains the configuration state of a respective switch within the fabric switch. Switch 700 maintains the state of the fabric switch, which is used to join other switches. Under such a scenario, communication ports 702 can include inter-switch communication channels for communication within a fabric switch. This inter-switch communication channel can be implemented via a regular communication port and based on any open or proprietary format (e.g., a TRILL or IP protocol).

In this example, switch 700 further includes a number of communication ports 702, a packet processor 710, a tunnel management module 730, a mapping module 730, a forwarding module 722, and a storage device 740. Tunnel management module 730 establishes an overlay tunnel with a remote switch based on an instruction for establishing the tunnel. Mapping module 730 maintains, in storage device 740, a mapping between a first tuple and a second tuple. The first tuple includes a MAC address and a VLAN identifier of a virtual machine. The second tuple includes the overlay tunnel and a network identifier.

Forwarding module 722 encapsulates a packet in a tunnel encapsulation header associated with the overlay tunnel based on the mapping. In some embodiments, packet processor 710 determines an egress port from communication ports 702 based on the destination address of the encapsulated packet and transmits the encapsulated packet via the egress port. Tunnel management module 730 terminates the overlay tunnel based on another instruction for terminating the tunnel.

Note that the above-mentioned modules can be implemented in hardware as well as in software. In one embodiment, these modules can be embodied in computer-executable instructions stored in a memory which is coupled to one or more processors in switch 700. When executed, these instructions cause the processor(s) to perform the aforementioned functions.

In summary, embodiments of the present invention provide a switch, a system, and a method for facilitating dynamic overlay tunnel orchestration. In one embodiment, the system detects the appearance of a first virtual machine running on a hypervisor of a first host machine coupled to a first switch in a network. The system identifies a VLAN associated with the first virtual machine and determines whether an overlay tunnel exists between the first switch and a second switch coupling a second virtual machine belonging to the first VLAN. If no such overlay tunnel exists, the system generates an instruction for the first and second switches to establish a first overlay tunnel between the first and second switches.

The methods and processes described herein can be embodied as code and/or data, which can be stored in a computer-readable non-transitory storage medium. When a computer system reads and executes the code and/or data stored on the computer-readable non-transitory storage medium, the computer system performs the methods and processes embodied as data structures and code and stored within the medium.

The methods and processes described herein can be executed by and/or included in hardware modules or apparatus. These modules or apparatus may include, but are not limited to, an application-specific integrated circuit (ASIC) chip, a field-programmable gate array (FPGA), a dedicated or shared processor that executes a particular software module or a piece of code at a particular time, and/or other programmable-logic devices now known or later developed. When the hardware modules or apparatus are activated, they perform the methods and processes included within them.

The foregoing descriptions of embodiments of the present invention have been presented only for purposes of illustration and description. They are not intended to be exhaustive or to limit this disclosure. Accordingly, many modifications and variations will be apparent to practitioners skilled in the art. The scope of the present invention is defined by the appended claims.

Claims

1. A computing system, comprising: a processor; anda memory storing instructions that when executed by the processor cause the system to perform a method, the method comprising: detecting appearance of a first virtual machine running on a hypervisor of a first host machine reachable via a first switch in a network of interconnected switches by obtaining information indicating the appearance of the first virtual machine from a notification message from a notification agent of the hypervisor on which the first virtual machine runs and detecting the appearance in accordance with the information indicating the appearance;identifying a first virtual local area network (VLAN) to which the first virtual machine belongs;identifying, in the network of interconnected switches, a second switch via which a second virtual machine belonging to the first VLAN is reachable;determining whether an overlay tunnel currently exists between the first switch and the second switch; andin response to determining that no overlay tunnel currently exists between the first and second switches, generating an instruction for the first and second switches to establish a first overlay tunnel between the first and second switches, and managing and maintaining the first overlay tunnel by the first switch and the second switch, including packet encapsulation performed at the first switch and second switch.

2. The computing system of claim 1, wherein the method further comprises allocating a network identifier for the first VLAN, wherein the network identifier indicates traffic belonging to the first VLAN in the first overlay tunnel.

3. The computing system of claim 2, wherein the method further comprises maintaining a mapping between a VLAN identifier of the first VLAN and a tuple comprising a tunnel identifier of the first overlay tunnel and the network identifier.

4. The computing system of claim 1, wherein the method further comprises refraining from instructing the first and second switches to establish the first overlay tunnel in response to determining that an overlay tunnel currently exists between the first and second switches.

5. The computing system of claim 1, wherein the method further comprises: detecting that the first virtual machine is no longer reachable via the first switch;determining whether the first overlay tunnel provides an overlay service to the first and second switches;in response to determining that the first overlay tunnel does not provide an overlay service, generating an instruction for the first and second switches to terminate the first overlay tunnel between the first and second switches.

6. The computing system of claim 5, wherein determining whether the first overlay tunnel provides an overlay service to the first and second switches comprises one or more of: determining whether the first overlay tunnel carries traffic belonging to a second VLAN; anddetermining whether a third virtual machine belonging to the first VLAN is reachable via the first switch.

7. The computing system of claim 5, wherein the first virtual machine has been removed from the first host machine based on one of: a migration of the first virtual machine from the first host machine; anda deletion of the first virtual machine from the first host machine.

8. A computer-executable method, comprising: detecting appearance of a first virtual machine running on a hypervisor of a first host machine reachable via a first switch in a network of interconnected switches by obtaining information indicating the appearance of the first virtual machine from a notification message from a notification agent of the hypervisor on which the first virtual machine runs and detecting the appearance in accordance with the information indicating the appearance;identifying a first virtual local area network (VLAN) to which the first virtual machine belongs;identifying, in the network of interconnected switches, a second switch via which a second virtual machine belonging to the first VLAN is reachable;determining whether an overlay tunnel currently exists between the first switch and the second switch; andin response to determining that no overlay tunnel currently exists between the first and second switches, generating an instruction for the first and second switches to establish a first overlay tunnel between the first and second switches, and managing and maintaining the first overlay tunnel by the first switch and the second switch, including packet encapsulation performed at the first switch and second switch.

9. The method of claim 8, further comprising allocating a network identifier for the first VLAN, wherein the network identifier indicates traffic belonging to the first VLAN in the first overlay tunnel.

10. The method of claim 9, further comprising maintaining a mapping between a VLAN identifier of the first VLAN and a tuple comprising a tunnel identifier of the first overlay tunnel and the network identifier.

11. The method of claim 8, further comprising refraining from instructing the first and second switches to establish the first overlay tunnel in response to determining that an overlay tunnel currently exists between the first and second switches.

12. The method of claim 8, further comprising: detecting that the first virtual machine is no longer reachable via the first switch;determining whether the first overlay tunnel provides an overlay service to the first and second switches;in response to determining that the first overlay tunnel does not provide an overlay service, generating an instruction for the first and second switches to terminate the first overlay tunnel between the first and second switches.

13. The method of claim 12, wherein determining whether the first overlay tunnel provides an overlay service to the first and second switches comprises one or more of: determining whether the first overlay tunnel carries traffic belonging to a second VLAN; anddetermining whether a third virtual machine belonging to the first VLAN is reachable via the first switch.

14. The method of claim 12, wherein the first virtual machine has been removed from the first host machine based on one of: a migration of the first virtual machine from the first host machine; anda deletion of the first virtual machine from the first host machine.

15. A switch, comprising: tunnel circuitry configured to establish an overlay tunnel with a second switch based on a first instruction identifying tunnel endpoint switches of the overlay tunnel, wherein a first virtual machine is reachable via the switch and a second virtual machine is reachable via the second switch, and wherein the first and second virtual machines belong to a first virtual local area network (VLAN), in response to information indicating an appearance of the first virtual machine running on a hypervisor of a first host machine from a notification message from a notification agent of the hypervisor on which the first virtual machine runs; andmapping circuitry configured to maintain, in a storage device, a mapping between a first tuple and a second tuple, wherein the first tuple comprises a media access control (MAC) address of the second virtual machine and a VLAN identifier of the first VLAN, wherein the second tuple comprises an tunnel identifier of the overlay tunnel and a network identifier, and wherein the network identifier indicates traffic belonging to the VLAN in the overlay tunnel, wherein the tunnel circuitry in cooperation with the second switch manage and maintain the overlay tunnel and provide packet encapsulation.

16. The switch of claim 15, wherein the first instruction further comprises the VLAN identifier, the overlay tunnel, and the network identifier.

17. The switch of claim 15, further comprising forwarding circuitry configured to encapsulate a packet in a tunnel encapsulation header associated with the overlay tunnel based on the mapping, wherein the tunnel encapsulation header includes the network identifier.

18. The switch of claim 15, wherein the tunnel circuitry is further configured to terminate the overlay tunnel with the second switch based on a second instruction that identifies tunnel endpoint switches for the overlay tunnel.