Detecting and resolving multicast traffic performance issues

Abstract

The subject disclosure relates to systems and methods for improving multicast traffic flows in a computer network. In some aspects, a method of the technology includes steps for receiving multicast traffic statistics from each of a plurality of switches in a computer network, aggregating the multicast traffic statistics into a time-series database, and identifying a low-performing multicast flow based on the time-series database. In some aspects, the method can include steps for automatically reconfiguring the computer network to improve the low-performing multicast flow. Systems and machine readable media are also provided.

Description

BACKGROUND

1. Technical Field

The subject technology relates systems and methods for identifying and correcting performance problems in multicast traffic flows. In particular, aspects of the technology provide solutions for moving (e.g., re-rooting) multicast trees to improve multicast performance.

2. Introduction

In some network configurations, several end user terminals, or hosts, may wish to receive the same data at the same time. This data can include anything from video or audio content, or software updates, to information about the network itself. While it would be possible to send this information simultaneously and individually to each host in the network, this would involve the transmission of replicated data throughout the network. Methods of multicasting data have therefore been developed in which data is transmitted through the network only to those destinations or hosts that have indicated a desire to receive the data, for example, by joining a corresponding multicast group. Generally, multicast data is replicated in the network only where the route to two destination hosts splits. Therefore, only one copy of the data is sent through the network until routes to the destination hosts diverge. Data is therefore sent through the network in a multicast tree, from which branches are formed as destination routes diverge.

BRIEF DESCRIPTION OF THE DRAWINGS

Certain features of the disclosed technology are set forth in the appended claims. However, the accompanying drawings, which are included to provide further understanding, illustrate certain aspects and together with the description explain the principles of the subject technology. In the drawings:

FIG. 1 illustrates an example network environment in which some aspects of the technology can be implemented.

FIG. 2A illustrates an example of a multicast flow (FTAG) and port matrix for a network switch, according to some aspects of the technology.

FIG. 2B illustrates a portion of a network fabric, including network devices such as leaf and spine switches, on which some aspects of the technology can be implemented.

FIG. 3 illustrates a flow chart of an exemplary algorithm that can be used to automatically resolve multicast traffic performance issues, according to some aspects of the technology.

FIG. 4 illustrates an example a network device configured to implement various aspects of the subject technology.

DETAILED DESCRIPTION

The detailed description set forth below is intended as a description of various configurations of the subject technology and is not intended to represent the only configurations in which the technology can be practiced. The appended drawings are incorporated herein and constitute a part of the detailed description. The detailed description includes specific details for the purpose of providing a more thorough understanding of the subject technology. However, it will be clear and apparent that the subject technology is not limited to the specific details set forth herein and may be practiced without these details. In some instances, structures and components are shown in block diagram form in order to avoid obscuring certain concepts of the technology.

Overview:

A method performed for implementing aspects of the technology can include steps for receiving multicast traffic statistics from each of a plurality of switches in a computer network, aggregating the multicast traffic statistics into a time-series database, and using the time-series database, automatically identifying a low-performing multicast flow. In some approaches, the method can further include steps for reconfiguring the network to improve the low-performing multicast flow, e.g., to reduce packet drop events, and improve traffic throughput. As discussed in further detail below, network reconfigurations can be performed automatically, e.g., by a network controller, or another suitable network process. Some reconfigurations can include the suppression of selected communication links (e.g., “links”) between network devices, such as link “pruning.” In some instances, network reconfiguration can involve the pruning of a given Forwarding Tag (“FTAG”), e.g., by removing a selected FTAG from a list of available FTAGs. Additionally, some network reconfigurations can involve the migration or re-rooting of an entire multicast tree, e.g., from a source switch to a new (destination) switch in the network.

Description:

An increasing amount of network traffic is the result of streaming media applications, such as Live TV. With the proliferation of media traffic, Internet Protocol (IP) multicast performance is rapidly becoming a focal point for cloud providers. When a media stream (such as a live video stream) experiences packet drops, the result is skipped frames, resulting in a degraded end-user experience. To ensure a high quality user experience, datacenters and other providers of media traffic need to proactively monitor their networks to mitigate packet drops.

It is often difficult for network administrators to fix multicast flow problems because conventional network deployments lack tools for identifying multicast issues. For example, in conventional configurations, network administrators must manually trace multicast traffic through the network and identify where drops are happening. This is typically performed by manually reading counters from each switch associated with a low performing multicast flow. Drop events must then be manually traced to pinpoint traffic bottlenecks. It is almost impossible to manually determine what path a multicast tree takes, what flows take those trees, and what trees are experiencing drops in a short enough time to make necessary network changes. Additionally, manually troubleshooting problems associated with multicast flows is not scalable, and therefore must be repeated every time the network experiences performance issues.

Aspects of the disclosed technology address the foregoing problems by providing systems and methods for identifying multicast drop events, and automatically pushing new network configurations to improve traffic flow quality. The technology can be implemented by a network controller, or other system enabled to configure/reconfigure network nodes and links necessary to re-rout multicast trees. In some aspects, the network controller is configured to automatically prune individual links associated with a multicast flow, automatically prune multicast trees associated with a multicast flow, and/or to re-root an entire multicast tree by transferring the root to a new destination switch, such as a new network spine that is determined to be suitable for the associated flow.

In some implementations, a network controller is configured to periodically receive traffic statistics from various switches e.g., spine switches, or other routing devices in the network fabric. Traffic statistics can be transmitted by each reporting switch, using a corresponding hardware offload engine, and can include various types of information about the switch's load and performance. By way of non-limiting example, traffic statistics can include information identifying one or more of: a corresponding switch/spine, port, multicast flow, flow bandwidth, port bandwidth, a number of transmitted or received FTAG packets per port, a total number of packets sent or received and/or packet drop counts.

It is understood that the various traffic statistics can be collected for unicast traffic (e.g., unicast statistics) and/or multicast traffic (e.g., multicast statistics), and transmitted by an offload engine. The traffic statistics are then aggregated by a monitoring device, such as a network controller, or other monitoring process, and stored into a time-indexed database.

The database can then be monitored to identify candidate multicast flows for which network reconfiguration may improve performance, i.e., for which link pruning, multicast (FTAG) pruning, or multicast root-transfer (re-rooting) could reduce drop events.

Selection of a given multicast flow can be performed based on a consideration for the overall impact that would be incurred by the network with the new network configuration. In some approaches, new configurations resulting in a smaller network impact are preferred over reconfigurations that would require major network changes and/or cause significant disruptions to other flows. For example, multicast flows with a greater percentage of packet drops and/or lower overall traffic bandwidth can be prioritized for reconfiguration. Conversely, multicast flows of greater size (bandwidth) may be of a lower priority, due to the potential for causing greater network disruptions. The various parameters used to identify/select network flows for improvement can vary depending on the desired implementation, and in some instances, can be configurable parameters, for example, that are set by a network administrator.

As used herein, link pruning refers to any reconfiguration that re-routes traffic away from an existing link between two network entities. Link pruning can be used to remove a link from the allowable FTAG link-set, so long as reachability is not affected. Multicast tree pruning refers to the elimination of one or more multicast trees from a source switch. Multicast tree pruning is typically performed in where multiple FTAG trees are available for broadcast, and a target multicast tree (FTAG) can be pruned without affecting reachability. In turn, multicast root-transfer (re-rooting), involves the movement of a multicast tree's root from a source switch (e.g., a source spine switch), to a destination switch, such as a new destination spine switch.

Due to a greater potential for traffic disruptions, multicast tree re-rooting may only be performed when link pruning and/or multicast tree pruning are insufficient to resolve performance issues. In some implementations, multicast re-rooting is not performed automatically, but rather, is presented as an option (e.g., to a user or network administrator), together with other information detailing the expected impact of the proposed reconfiguration. This information can include details about the new configuration to be pushed, identification of one or more other links, or potentially affected traffic flows, and/or an identification of one or more network devices to be effected.

FIG. 1 illustrates a diagram of an example network environment 100 in which multicast monitoring can be performed. Fabric 112 can represent the underlay (i.e., the physical network) of environment 100. Fabric 112 includes spine switches 1-N (102_{A N}) (collectively “102”) and leaf switches 1-N (104_A-N) (collectively “104”). Leaf switches 104 can reside at the edge of fabric 112, and can represent the physical network edges. Leaf switches 104 can be, for example, top-of-rack (“ToR”) switches, aggregation switches, gateways, ingress and/or egress switches, provider edge devices, and/or any other type of routing or switching device.

Leaf switches 104 can be responsible for routing and/or bridging tenant or endpoint packets and applying network policies. Spine 102 can perform switching and routing within fabric 112. Thus, network connectivity in fabric 112 can flow from spine switches 102 to leaf switches 104, and vice versa.

Leaf switches 104 can provide servers 1-4 (106_A-D) (collectively “106”), hypervisors 1-4 (108_A-108_D) (collectively “108”), virtual machines (VMs) 1-4 (110_A-110_D) (collectively “110”). For example, leaf switches 104 can encapsulate and decapsulate packets to and from servers 106 in order to enable communications throughout environment 100. Leaf switches 104 can also connect other network-capable device(s) or network(s), such as a firewall, a database, a server, etc., to the fabric 112. Leaf switches 104 can also provide any other servers, resources, endpoints, external networks, VMs, services, tenants, or workloads with access to fabric 112.

As discussed in further detail with respect to FIG. 2B below, leaf switches 104 and spine switches 102 can be used to carry multicast flows, for example, to/from one or more hosts in the network (not illustrated).

FIG. 2A illustrates an example of a multicast flow (FTAG) and port matrix 200, that represents FTAG/port associations for a given network switch. Port matrix 200 can be used to facilitate link pruning decisions. The FTAG/port matrix can be compiled and stored at the associated switch and/or a centralized network controller that is coupled to the switch. Port matrix 200 can be used by the network controller to identify candidate links that can be pruned to improve at least one multicast flow. As discussed in detail below, in some instances, link pruning decisions should be based on determinations of host reachability and bandwidth reallocation, i.e., the availability of one or more other links to absorb bandwidth for a pruned link. That is, candidate links can be ranked higher for pruning where host reachability is not impacted, and/or where smaller amounts of traffic are to be redirected in the network.

In some aspects, reachability link pruning decisions can also be pre-conditioned on network reachability. That is, candidate links can only be pruned if it is first determined that reachability for network traffic flowing over those links will not be impacted, i.e., reduced.

Link utilization statistics for each FTAG/port pair can be compared to determine packet drop percentages and bandwidth statistics for each link. In some implementations, bandwidth statistics can include combined measures for multicast traffic and/or unicast traffic; however, in some preferred implementations, only multicast traffic may be considered. The FTAG, port pair with the highest drop percentage, and/or lowest total bandwidth on that FTAG can be selected for link pruning. By way of example, matrix 200 illustrates a configuration in which each port has FTAGs A, B, or C. If FTAG B sends 1 million packets, and experiences 500 packet drops, whereas FTAG A sends 100 thousand packets and experiences 500 packet drops, then FTAG A, having the higher drop percentage, may be prioritized for pruning over FTAG B.

Once a link associated with a specific FTAG is selected for pruning, it is determined if any hosts connected to the FTAG will no longer be reachable if the link is pruned, i.e., to determine if the selected link provides a redundant path. Further to the example illustrated in FIG. 2A, FTAG A includes both ports 0 and 1 as fanout links. If any destination (host) subscribed to FTAG A is reachable through both ports 0 and 1, then port 0 can be pruned from FTAG A, as the traffic can be routed through port 1, without compromising reachability.

In some aspects, link pruning can be skipped if an alternative link is unavailable, or unable to absorb the traffic bandwidth from the pruned link. That is, link pruning decisions can also take into consideration whether the required increase in bandwidth for one or more alternative links exceeds an allowed bandwidth for those links. Further to the above example, pruning FTAG A at port 0 would cause the traffic to be routed through port 1, increasing the bandwidth throughput on port 1. As such, the increased traffic load on port 1 should not result in packet drops.

In some aspects, if it is determined that a bandwidth threshold for a new target port/link would be exceeded, alternative links can be considered for pruning. Alternatively, if no candidate FTAG links can be pruned, then determinations can be made regarding multicast tree pruning, as discussed in further detail with respect to network environment 201, below.

FIG. 2B illustrates an example of a partial network environment 201, including network devices on which some aspects of the technology can be implemented. Environment 201 includes spine switches (Spine 0 and Spine 1), each connected to leaf switches (Leaf 0, and Leaf 1). It is understood that environment 201 only represents a portion of a network fabric, for example, as described with respect to network 100 in FIG. 1, above. In the illustrated example, spine and leaf switches are configured for carrying multicast flows originating from at least one host device connected to Leaf 0 and/or Leaf 1 (not illustrated), and for providing the multicast traffic to recipient hosts (not illustrated).

In implementations wherein the source switch (e.g., Spine 0 or Spine 1) can be configured to choose multiple FTAG trees for broadcast, selected FTAG trees can be pruned to reduce the load corresponding with that FTAG. In the example of FIG. 2B, if Spine 0 experiences FTAG A drops on a particular downlink port, bandwidth contributions for each link into Spine 0 (e.g., A0, and A1) can be computed to determine which provides the smallest bandwidth contribution. If link A0 is determined to have the smallest contribution that can also account for the packet drops on the downlink, link A0 may be more highly ranked for potential pruning than link A1. Subsequently, each ranked link can be examined to determine if the associated FTAG tree is part of an equal-cost multi-path (ECMP) set. In the illustrated example, Leaf 0 to Spine 0 has an ECMP set of FTAGs, i.e., FTAG A and FTAG B. If both FTAG A and FTAG B provide the same subscriber reachability, then either can be considered valid options for sending traffic, and thus FTAG A on Leaf 0 could be a candidate for pruning.

In order to complete FTAG pruning, it is first determined if alternate links have the capacity to absorb the additional traffic. Further to the above example, to prune FTAG A from Leaf 0, the link from Leaf 0 to Spine 1 (link B0) would need the capacity to absorb the traffic from link A0. If pruning the selected FTAG would result in packet drop mitigation, then FTAG A can be pruned, e.g., from the source ECMP list on Leaf 0. If flow performance would not be improved by pruning FTAG A, additional analysis/ranking of links can be performed, e.g., by re-analyzing reachability for the various FTAGs.

In some implementations, it may be determined that there are no suitable solutions for FTAG pruning. In such instances, multicast FTAG re-rooting can be considered. However, before FTAG movement is performed, suitability of FTAG re-rooting is considered, for example, by the controller (or other network monitoring process), on a spine-by-spine basis to determine if re-rooting in a new spine would improve flow performance. For example, if FTAG A is rooted in Spine 0 and needs to be moved, candidate spines could include Spine 1, as well as one or more other spines (e.g., Spine 2, not illustrated). If Spine 1 provides similar subscriber reachability as Spine 0, but Spine 2 does not, then only Spine 1 may be considered as a destination candidate for FTAG A.

Other parameters can also be considered. For example, before multicast re-rooting is performed, various ingress and egress ports of the candidate spine switch can be analyzed to determine if the added multicast bandwidth can be accommodated. By way of example, ingress and egress ports of Spine 1 can be analyzed to see if each can accommodate additional bandwidth resulting from traffic on links A0 and A1, should FTAG A be re-rooted in Spine 1. If the candidate destination spine is capable of absorbing the added traffic from FTAG A, re-rooting can be performed. Alternatively, no re-rooting is attempted, and a message or indication may be delivered to the network administrator, e.g., via the network controller, to indicate that multicast flow performance issues could not be resolved through network reconfiguration.

Although re-rooting can be performed automatically, because multicast re-rooting requires significant network configuration changes, in some aspects, re-rooting may be provided as an option, for example, to the network administrator before any network changes are implemented. In such instances, the administrator can be provided an alert indicating the new configuration to be pushed to move the multicast tree.

FIG. 3 illustrates a flow chart of an example process 300 used to resolve multicast traffic performance issues, according to some aspects of the technology. Process 300 begins with step 302 in which multicast statistics are received from multiple switches in a computer network. The multicast statistics can be received by a network controller, or other monitoring hardware and/or software. Statistics can be provided by an offload engine for a respective switch/router, and can include various types of metrics relating to current or historic device performance. For example, multicast statistics can indicate the frequency of packet drop events, or bandwidth statistics for particular flows, e.g., on a port-by-port basis.

In step 304, the multicast statistics are aggregated (e.g., by the network controller), into a time-series database. The time-series database can provide a snapshot of current and historic flow performance, e.g., for various FTAGs, across different leaf/spine switches in the network fabric. Subsequently, at step 306, low performing multicast flows are identified through analysis of the time-series database. As discussed above with respect to FIG. 2A, flows experiencing performance degradation can be identified for each switch/router on a port-by-port basis. Flow statistics for each FTAG/port pair can be aggregated for multiple switches/routers to determine overall performance for the multicast flow. Additionally, several different types of flow statistics can be considered when prioritizing multicast flows. For example, a flow can be selected/prioritized for further inspection based on a ratio of packet drop events to flow size (bandwidth). That is, flows with greater proportions of packet drops can be prioritized. Additionally, smaller multicast flows may be given precedence since redirection of smaller traffic volumes can be less disruptive to overall network performance.

Once a flow has been identified, different network reconfiguration options can be considered to improve flow performance. As discussed above, link pruning may be considered first, as link pruning can require the least impactful network configuration changes. If link pruning is determined to likely improve flow performance, without affecting host reachability, a new network configuration can be automatically pushed to redirect traffic around the pruned link/s. In some aspects, the new network configuration may first be provided to a network administrator or other user, for example, to give the administrator an opportunity to approve the new configuration before it is pushed. Alternatively, if link pruning is not possible, or determined to not improve flow performance, multicast pruning (e.g., FTAG pruning) can be considered.

In instances where multicast pruning is determined to improve flow performance, a new network configuration can be automatically pushed, e.g., to prune one or more identified FTAGs necessary to improve flow performance. In some approaches, multicast pruning may be presented as an option to a network administrator, e.g., via a network controller, so that the network administrator can approve the new network configuration before changes are pushed.

In aspects wherein multicast pruning is determined to not be a viable option, multicast re-rooting is considered. Determinations regarding whether multicast re-rooting can be performed can depend on the availability and performance metrics of a viable candidate switch (e.g., spine switch) in which the multicast flow can be re-rooted. As discussed above, subscriber reachability by the candidate destination is determined. Candidate destination switches offering full, or substantially complete, subscriber reachability can be preferred over switches with limited or less complete reachability. Because multicast re-rooting can involve substantial network reconfiguration, re-rooting may be presented as an option to the network administrator, including various details of the overall network impact, before configuration changes get pushed to the network.

FIG. 4 illustrates an example network device 410 that can be used to implement a spine and/or leaf router, as discussed above. Network device 410 includes a master central processing unit (CPU) 462, interfaces 468, and a bus 415 (e.g., a PCI bus). When acting under the control of appropriate software or firmware, the CPU 462 is responsible for executing packet management, error detection, and/or routing functions. The CPU 462 preferably accomplishes all these functions under the control of software including an operating system and any appropriate applications software. CPU 462 may include one or more processors 463 such as processors from the Intel, ARM, MIPS or Motorola family of microprocessors. In an alternative embodiment, processor 463 is specially designed hardware for controlling the operations of router 410. In a specific embodiment, a memory 461 (such as non-volatile RAM and/or ROM) also forms part of CPU 462. However, there are many different ways in which memory could be coupled to the system.

Interfaces 468 can be provided as interface cards (sometimes referred to as “line cards”). Generally, they control the sending and receiving of data packets over the network and sometimes support other peripherals used with the router 410. Among the interfaces that may be provided are Ethernet interfaces, frame relay interfaces, cable interfaces, DSL interfaces, token ring interfaces, and the like. In addition, various very high-speed interfaces may be provided such as fast token ring interfaces, wireless interfaces, Ethernet interfaces, Gigabit Ethernet interfaces, ATM interfaces, HSSI interfaces, POS interfaces, FDDI interfaces and the like. Generally, these interfaces may include ports appropriate for communication with the appropriate media. In some cases, they may also include an independent processor and, in some instances, volatile RAM. The independent processors may control such communications intensive tasks as packet switching, media control and management. By providing separate processors for the communications intensive tasks, these interfaces allow the master microprocessor 462 to efficiently perform routing computations, network diagnostics, security functions, etc.

Although the system shown in FIG. 4 is one specific network device of the present invention, it is by no means the only network device architecture on which the present invention can be implemented. For example, an architecture having a single processor that handles communications as well as routing computations, etc. is often used. Further, other types of interfaces and media could also be used with the router.

Regardless of the network device's configuration, it may employ one or more non-transitory memories or memory modules (including memory 461) configured to store program instructions for general-purpose network operations and mechanisms necessary to implement the network reconfiguration methods discussed above. For example, memory 461 can include a non-transitory computer-readable medium that includes instructions for causing CPU 462 to execute operations for receiving traffic statistics from each of a plurality of switches in a computer network, aggregating the statistics into a time-series database, and automatically identifying a low-performing multicast flow, based on the time-series database. In some implementations, memory 461 can further include instructions for reconfiguring the computer network to improve the low-performing multicast flow.

It is understood that any specific order or hierarchy of steps in the processes disclosed is an illustration of exemplary approaches. Based upon design preferences, it is understood that the specific order or hierarchy of steps in the processes may be rearranged, or that only a portion of the illustrated steps be performed. Some of the steps may be performed simultaneously. For example, in certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

The previous description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects. Thus, the claims are not intended to be limited to the aspects shown herein, but are to be accorded the full scope consistent with the language claims, wherein reference to an element in the singular is not intended to mean “one and only one” unless specifically so stated, but rather “one or more.”

A phrase such as an “aspect” does not imply that such aspect is essential to the subject technology or that such aspect applies to all configurations of the subject technology. A disclosure relating to an aspect may apply to all configurations, or one or more configurations. A phrase such as an aspect may refer to one or more aspects and vice versa. A phrase such as a “configuration” does not imply that such configuration is essential to the subject technology or that such configuration applies to all configurations of the subject technology. A disclosure relating to a configuration may apply to all configurations, or one or more configurations. A phrase such as a configuration may refer to one or more configurations and vice versa.

The word “exemplary” is used herein to mean “serving as an example or illustration.” Any aspect or design described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other aspects or designs.

Claims

1. A computer-implemented method comprising: receiving, at a network controller, multicast traffic statistics from each of a plurality of switches in a computer network;aggregating, by the network controller, the multicast traffic statistics into a time-series database;automatically identifying, by the network controller, a low-performing multicast flow, based on the time-series database, wherein the low-performing multicast flow is associated with a source switch; andreconfiguring the computer network to improve the low-performing multicast flow;wherein identifying the low-performing multicast flow further comprises computing a total packet throughput for one or more forwarding tag (FTAG) and port pairs.

2. The computer-implemented method of claim 1, wherein reconfiguring the computer network further comprises: determining, by the network controller, if a multicast root corresponding with the low-performing multicast flow can be transferred to a destination switch in the computer network; andreconfiguring the source switch and the destination switch to transfer the multicast root to the destination switch.

3. The computer-implemented method of claim 2, wherein reconfiguring the source switch and the destination switch further comprises: providing an alert to a network administrator, via the network controller, the alert indicating a recommended network configuration to improve the low-performing multicast flow; andreceiving a response from the network administrator, the response indicating either an approval or rejection of the recommended network configuration.

4. The computer-implemented method of claim 1, wherein reconfiguring the computer network further comprises: automatically pruning one or more links associated with the low-performing multicast flow.

5. The computer-implemented method of claim 1, wherein reconfiguring the computer network further comprises: automatically pruning a multicast tree associated with the low-performing multicast flow.

6. A network controller for improving multicast traffic flows, comprising: one or more processors; anda non-transitory computer-readable medium comprising instructions stored therein, which when executed by the processors, cause the processors to perform operations comprising: receiving multicast traffic statistics from each of a plurality of switches in a computer network;aggregating the multicast traffic statistics into a time-series database;automatically identifying a low-performing multicast flow, based on the time-series database, wherein the low-performing multicast flow is associated with a source switch; andreconfiguring the computer network to improve the low-performing multicast flow;wherein identifying the low-performing multicast flow comprises computing a total packet throughput for one or more forwarding tag (FTAG) and port pairs.

7. The network controller of claim 6, wherein reconfiguring the computer network further comprises: determining if a multicast root corresponding with the low-performing multicast flow can be transferred to a destination switch in the computer network; andreconfiguring the source switch and the destination switch to transfer the multicast root to the destination switch.

8. The network controller of claim 7, wherein reconfiguring the source switch and the destination switch further comprises: providing an alert to a network administrator the alert indicating a recommended network configuration to improve the low-performing multicast flow; andreceiving a response from the network administrator, the response indicating either an approval or rejection of the recommended network configuration.

9. The network controller of claim 6, wherein reconfiguring the computer network further comprises: automatically pruning one or more links associated with the low-performing multicast flow.

10. The network controller of claim 6, wherein reconfiguring the computer network further comprises: automatically pruning a multicast tree associated with the low-performing multicast flow.

11. A non-transitory computer-readable storage medium comprising instructions stored therein, which when executed by one or more processors, cause the processors to perform operations comprising: receiving multicast traffic statistics from each of a plurality of switches in a computer network;aggregating the multicast traffic statistics into a time-series database;automatically identifying a low-performing multicast flow, based on the time-series database, wherein the low-performing multicast flow is associated with a source switch; andreconfiguring the computer network to improve the low-performing multicast flow;wherein identifying the low-performing multicast flow further comprises computing a total packet throughput for one or more forwarding tag (FTAG) and port pairs.

12. The non-transitory computer-readable storage medium of claim 11, wherein reconfiguring the computer network further comprises: determining if a multicast root corresponding with the low-performing multicast flow can be transferred to a destination switch in the computer network; andreconfiguring the source switch and the destination switch to transfer the multicast root to the destination switch.

13. The non-transitory computer-readable storage medium of claim 12, wherein reconfiguring the source switch and the destination switch further comprises: providing an alert to a network administrator, the alert indicating a recommended network configuration to improve the low-performing multicast flow; andreceiving a response from the network administrator, the response indicating either an approval or rejection of the recommended network configuration.

14. The non-transitory computer-readable storage medium of claim 11, wherein reconfiguring the computer network further comprises: automatically pruning one or more links associated with the low-performing multicast flow.

15. The non-transitory computer-readable storage medium of claim 11, wherein reconfiguring the computer network further comprises: automatically pruning a multicast tree associated with the low-performing multicast flow.