WO2020119682A1 - Procédé de partage de charge, entité de plan de commande et répéteur - Google Patents

Procédé de partage de charge, entité de plan de commande et répéteur Download PDF

Info

Publication number
WO2020119682A1
WO2020119682A1 PCT/CN2019/124319 CN2019124319W WO2020119682A1 WO 2020119682 A1 WO2020119682 A1 WO 2020119682A1 CN 2019124319 W CN2019124319 W CN 2019124319W WO 2020119682 A1 WO2020119682 A1 WO 2020119682A1
Authority
WO
WIPO (PCT)
Prior art keywords
bandwidth
equal
repeater
scheduling unit
management component
Prior art date
Application number
PCT/CN2019/124319
Other languages
English (en)
Chinese (zh)
Inventor
郑坤
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2020119682A1 publication Critical patent/WO2020119682A1/fr

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/52—Queue scheduling by attributing bandwidth to queues
    • H04L47/522—Dynamic queue service slot or variable bandwidth allocation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/08—Load balancing or load distribution
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/12—Avoiding congestion; Recovering from congestion
    • H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering

Definitions

  • Embodiments of the present disclosure relate to, but are not limited to, a method of load sharing, a control plane entity, a cluster management component, a load sharer, a repeater management component, a repeater, and a computer-readable storage medium.
  • BRAS Broadband Remote Access Server
  • BRAS Broadband Remote Access Server
  • BRAS Broadband Remote Access Server
  • BRAS is an access gateway for broadband network applications. It is a bridge between the broadband access network and the backbone network, and provides basic access methods and management functions of the broadband access network. It is located at the edge of the network, provides broadband access services, realizes the convergence and forwarding of multiple services, and can meet the requirements of different users for transmission capacity and bandwidth utilization, so it is the core equipment for broadband users to access.
  • BRAS first authenticates the user who has access. The authentication allows the user to access the broadband network, and the authentication fails to reject the user's access.
  • FIG. 1 is a diagram of the architecture of the forwarding plane and the control plane.
  • the forwarding plane consists of multiple transponders, which are controlled by the control plane.
  • the control plane includes LB (load balance), multiple service processing instances, cluster management components, and repeater management components.
  • LB is the entrance to the control plane. All reported packets and reported data/events go through the LB. The uploaded packets go through the LB to implement the load sharing strategy, and then select the business processing instance according to the load sharing strategy and forward the packets to the instance Perform control business processing. Since multiple transponders managed by the forwarding plane have a relatively large number of reported packets, LB is required to have strong IO (Input/Output, input/output) and CPU (Central Processing Unit, central processing unit) computing power.
  • the repeater management component manages the repeater, and the data/events reported by the repeater are also directly sent to the repeater management component through LB.
  • the entire control plane is a cluster system, and the management of the system is the responsibility of the cluster management component.
  • the services on the transponders on the forwarding plane are not uniform. Some transponders have a large business volume and some have a small business volume. The business packet traffic of the business volume is large. The LB sent to the control plane will occupy more bandwidth. The packet traffic sent by other transponders cannot be forwarded, which affects the service processing of other transponders. What's more, malicious attackers carried out starvation attacks on the system, sending a large number of illegal messages, impacting the control plane and starving to work properly.
  • Embodiments of the present disclosure provide a method of load sharing, a control plane entity, a cluster management component, a load sharer, a repeater management component, a repeater, and a computer-readable storage medium to reasonably allocate repeater bandwidth.
  • An embodiment of the present disclosure provides a method of load sharing, including: the control plane divides the total bandwidth of the packet into multiple equal bandwidths; the control plane allocates one or more equal bandwidths to the repeater, Associating the allocated one or more equal bandwidths with the repeater, and associating the allocated one or more equal bandwidths with the service processing instance; the control plane receives The message sent by the repeater forwards the message to the corresponding service processing instance through the equal bandwidth associated with the repeater.
  • An embodiment of the present disclosure also provides a load sharing method, which includes: a cluster management component divides the total bandwidth of a message carried by the load balancer LB into multiple equal bandwidths; the cluster management component allocates one or more to the repeater Equal parts of the bandwidth, the one or more equal parts of the bandwidth is associated with the repeater, and the assigned one or more equal parts of the bandwidth is associated with the service processing instance, so that The message sent by the repeater to the LB is forwarded to the corresponding service processing instance through the equal bandwidth associated with the repeater.
  • An embodiment of the present disclosure also provides a load sharing method, including: LB receiving a message sent by a repeater; wherein the repeater is associated with one or more equal bandwidths, and the one or more shares The equal bandwidth of is associated with the business processing instance; the LB forwards the message to the corresponding business processing instance through the equal bandwidth associated with the forwarder.
  • An embodiment of the present disclosure also provides a load sharing method, which includes: the repeater management component determines one or more equal bandwidths allocated by the cluster management component to the repeater; the repeater management component The one or more equal bandwidths are associated with the repeater, so that the message sent by the repeater to the LB is forwarded to the corresponding service processing instance through the equal bandwidth associated with the repeater.
  • An embodiment of the present disclosure also provides a load sharing method, which includes: a forwarder obtains an encapsulation table of a forwarded message; wherein the encapsulation table of the forwarded message indicates that the forwarder and one or more equal bandwidths Correlation; the repeater sends the message to the LB according to the encapsulation table of the uploaded message, so that the message is forwarded to the corresponding business process through the equal bandwidth associated with the repeater Examples.
  • An embodiment of the present disclosure also provides a control plane entity, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the method for implementing the load sharing when the processor executes the program .
  • An embodiment of the present disclosure also provides a cluster management component, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the method for implementing the load sharing when the processor executes the program .
  • An embodiment of the present disclosure also provides a load sharer LB, including: a memory, a processor, and a computer program stored on the memory and executable on the processor.
  • the processor implements the program to implement the load sharing method.
  • An embodiment of the present disclosure also provides a repeater management component, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor realizes the load sharing when the program is executed method.
  • An embodiment of the present disclosure also provides a transponder, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the program to implement the load sharing method.
  • Embodiments of the present disclosure also provide a computer-readable storage medium that stores computer-executable instructions that are used to perform the load sharing method.
  • Embodiments of the present disclosure include: dividing the total bandwidth of the packet into multiple equal bandwidths; allocating one or more equal bandwidths to the repeater, and dividing the allocated one or more equal bandwidths and Associating the repeater, and associating the allocated one or more equal bandwidths with the service processing instance; receiving the message sent by the repeater, and passing the message to the forwarder
  • the equal bandwidth associated with the server is forwarded to the corresponding service processing instance.
  • the packet traffic on the repeater reaching the control plane can only occupy the bandwidth value of all the allocated bandwidth, and the packet traffic on the equal bandwidth can only be forwarded to the associated business processing instance to avoid impact Traffic from other transponders.
  • Figure 1 is the architecture diagram of the software-defined BRAS forwarding plane and control plane
  • FIG. 2 is a system architecture diagram of an embodiment of the present disclosure
  • FIG. 3 is a flowchart of a load sharing method (applied to a control plane) of an embodiment of the present disclosure
  • step 103 is a flowchart of step 103 of an embodiment of the present disclosure.
  • FIG. 5 is a schematic diagram of a packet forwarding process carrying a mapping identifier according to an embodiment of the present disclosure
  • FIG. 6 is a flowchart of a load sharing method (applied to a control plane) according to another embodiment of the present disclosure
  • FIG. 7 is a schematic diagram of dynamic adjustment of LB forwarded message forwarding according to an embodiment of the present disclosure
  • FIG. 8 is a schematic diagram of three stages of system startup and operation according to an embodiment of the present disclosure.
  • FIG. 9 is a flowchart of a load sharing method (applied to a cluster management component) of an embodiment of the present disclosure.
  • FIG. 10 is a schematic diagram of the composition of a cluster management component of an embodiment of the present disclosure.
  • FIG. 11 is a flowchart of a load sharing method (applied to LB) of an embodiment of the present disclosure
  • FIG. 12 is a flowchart of a load sharing method (applied to LB) of another embodiment of the present disclosure.
  • step 802 is a flowchart of step 802 according to an embodiment of the present disclosure.
  • 15 is a flowchart of a load sharing method (applied to a repeater management component) of an embodiment of the present disclosure
  • 16 is a schematic diagram of the composition of the repeater management component of the embodiment of the present disclosure.
  • 17 is a schematic diagram of the composition of a business processing example of an embodiment of the present disclosure.
  • FIG. 19 is a schematic diagram of the composition of a repeater according to an embodiment of the present disclosure.
  • 21 is a schematic flow chart of forwarding a forwarded message to a service processing unit after allocating an equal portion of bandwidth to an application example of the present disclosure
  • 22 is a schematic diagram of a process of automatically adjusting the bandwidth of a repeater when a performance problem is detected in an application example of the present disclosure
  • FIG. 23 is a schematic flowchart of an example of a business processing instance for automatically adjusting packet loss when a service processing instance interface of an application instance of the present disclosure is shown;
  • FIG. 24 is a schematic flowchart of an alternative application package type of the disclosure.
  • FIG. 25 is a schematic flow chart of the application example of the present disclosure where the package type is inserted into the tail;
  • FIG. 26 is a schematic flowchart of mapping a mapping identifier to a field of another message in an application example of the present disclosure
  • 27 is a schematic diagram of PPPoE access networking of two transponders in an application example of the present disclosure
  • FIG. 28 is a schematic flow chart of the flow of packets sent by a transponder in an application example of the present disclosure is limited within a range;
  • FIG. 29 is a schematic diagram of the composition of a control plane entity of an embodiment of the present disclosure.
  • FIG. 30 is a schematic diagram of the composition of a cluster management component of an embodiment of the present disclosure.
  • FIG. 31 is a schematic diagram of the composition of the LB embodiment of the present disclosure.
  • FIG. 33 is a schematic diagram of the composition of a repeater according to an embodiment of the present disclosure.
  • the bandwidth on the LB is divided into equal bandwidths according to a certain granularity, and one or more equal bandwidths are related to the repeater And one or more equal bandwidths are associated with the service processing instance.
  • the packet traffic on the forwarder reaches the LB, it can only occupy the allocated equal bandwidth, and the packet traffic on the equal bandwidth can only be forwarded to the association Example of business processing.
  • the load sharing method of the embodiment of the present disclosure which is applied to the control plane, includes:
  • Step 101 Divide the total bandwidth of the message into multiple equal bandwidths.
  • the total bandwidth of the message is divided into multiple equal bandwidths according to a preset granularity.
  • the cluster management component divides the total bandwidth of the LB into multiple equal bandwidths according to the granularity defined by the user policy, and assigns an ID (Identifier) to each equal bandwidth.
  • Each ID of the equal bandwidth has a unique ID Sex cannot be the same as other equal bandwidths.
  • the method further includes:
  • a scheduling unit is generated according to the equal bandwidth, each equal bandwidth corresponds to a scheduling unit, and the speed limit of the bandwidth is performed by the scheduling unit, and the bandwidth value of the speed limit of the scheduling unit is the bandwidth value of the corresponding equal bandwidth.
  • the scheduling unit is located in the LB, and specific implementation includes, but is not limited to: software implementation of the general server, programmable hardware implementation, and hardware circuit implementation. These implementations are specifically related to business scenarios, performance, and cost considerations.
  • Step 102 allocate one or more equal bandwidths to the repeater, associate the allocated one or more equal bandwidths with the repeater, and allocate the one or more equal bandwidth An equal portion of the bandwidth is associated with the business processing instance.
  • allocating one or more equal bandwidths to the repeater includes:
  • the bandwidth requirements of the repeater determine the minimum number of equal bandwidths required by the repeater, and allocate the minimum number of equal bandwidths to the repeater.
  • the repeater management component analyzes and estimates the repeater bandwidth requirement of the repeater, and notifies the cluster management component of the repeater bandwidth requirement.
  • the cluster management component calculates the minimum number of equal bandwidths, the maximum number of equal bandwidths, and the default equal bandwidth bandwidth required by the repeaters according to the bandwidth requirements of the forwarders, and then allocates the minimum equal bandwidth and divides the allocated equal bandwidth
  • the ID and equal bandwidth weights inform the repeater management component.
  • the associating the allocated one or more equal bandwidths with the repeater includes:
  • the repeater management component receives the equal bandwidth ID and the equal bandwidth weight allocated to the repeater, and generates an encapsulation table of the transmitted message.
  • the generating an encapsulation table of the transmitted packet according to the equal bandwidth identifier and the equal bandwidth bandwidth includes:
  • the mapping identifier is determined according to the identifier of the equal bandwidth, and the encapsulation table of the transmitted message is generated according to the mapping identifier and the weight of the equal bandwidth.
  • the upload packet encapsulation table specifies the field carried in the upload packet. This field identifies an equal bandwidth.
  • the equal bandwidth ID is used to generate this field according to a mapping policy (MP), which is called mapping. Identification (mapping identity, referred to as MI).
  • the mapping strategy may include but is not limited to the following strategies:
  • the first mapping strategy use the identifier of the equal bandwidth as the mapping identifier.
  • This mapping strategy encapsulates the equal bandwidth ID directly in the message.
  • MI is the equal bandwidth ID.
  • the second mapping strategy mapping the equal bandwidth identifier to a preset packet identifier, and using the preset packet identifier as the mapping identifier.
  • This mapping strategy maps equal bandwidth IDs into standard message fields, such as VLAN ID (Virtual Local Area Network Identifier), MPLS label (Multiprotocol Label Switching label, multi-protocol label switching label), tunnel ID , VxLAN, VNI (Virtual, Extensible, Network Identifier, etc.).
  • VLAN ID Virtual Local Area Network Identifier
  • MPLS label Multiprotocol Label Switching label, multi-protocol label switching label
  • tunnel ID e.g., VxLAN, VNI (Virtual, Extensible, Network Identifier, etc.
  • the encapsulation table of the uploaded message also carries an encapsulation type (encapsulation type, ET for short).
  • the encapsulation table of the forwarded message also specifies the encapsulation type.
  • the encapsulation type is used to indicate the protocol that encapsulates the mapping identifier, that is, which protocol is used to extend the encapsulation bandwidth ID.
  • the encapsulation type is used to indicate that the mapping identifier is encapsulated by insertion or replacement.
  • the mapping identifier is inserted according to the standard protocol.
  • the encapsulation type specifies whether the mapping identifies whether the encapsulation is inserted or replaced.
  • the inserting means that when the uploaded packet already contains the standard packet field defined by the mapping strategy before encapsulation, the mapping identifier is also inserted.
  • the package type is also used to indicate the order of insertion. The order refers to whether it is inserted in the head or the tail.
  • the replacement means that when the uploaded message already contains the message field defined by the mapping policy before encapsulation, the contained message field is deleted and replaced with the mapping identifier. Since the fields of some standard protocols can only appear once in the message, the encapsulation type of this mapping identification can only be replaced.
  • the mapping strategy and the encapsulation type can be specified by manual configuration.
  • the cluster management component and the repeater management component are notified.
  • the cluster management component notifies the LB of the mapping strategy and the encapsulation type. Since multiple equal bandwidth IDs may be allocated on the repeater, the selection weight of the mapping identification encapsulation is specified in the encapsulation table of the transmitted packet, and the selection weight of the mapping identification encapsulation is the weight of the equal bandwidth.
  • the repeater selects a mapping identifier for the transmitted message based on the equal bandwidth weight.
  • the selected methods include but are not limited to: polling, hashing and other methods.
  • the forwarder management component notifies the forwarder of the generated packet encapsulation table for forwarding.
  • the forwarder receives the entry and saves it.
  • the method further includes:
  • a scheduling unit query entry is generated according to the corresponding relationship between the equal bandwidth associated with the repeater and the mapping identifier, and the corresponding relationship between the equal bandwidth and the scheduling unit, and the scheduling unit query entry constitutes the scheduling unit query table.
  • the repeater management component informs the cluster management component of the correspondence between the equal bandwidth ID of each repeater and the mapping identifier.
  • the cluster management component queries the scheduling unit according to the equal bandwidth ID to generate the scheduling unit query table entry.
  • the keyword is the mapping identifier
  • the query result is the scheduling unit.
  • the scheduling unit query entry is notified to LB, and the LB receives the scheduling unit query entry and adds it to the scheduling unit query table.
  • the associating the allocated one or more equal bandwidths with the service processing instance includes: according to the correspondence between the equal bandwidth associated with the repeater and the mapping identifier, and the mapping The correspondence between the identifier and the business processing instance generates a business processing instance forwarding table.
  • the cluster management component generates a business processing instance forwarding table
  • the key of the entry is a mapping identifier
  • the query result is a business processing instance.
  • the processing capability of a business processing instance is relatively strong, so initially, multiple transponders can correspond to one business processing instance. That is to say, there is a case where different mapping identifiers query the same business processing instance for an entry in the business processing instance forwarding table.
  • Step 103 Receive a message sent by the repeater, and forward the message to the corresponding service processing instance through the equal bandwidth associated with the repeater.
  • the packet traffic on the repeater reaching the control plane can only occupy the bandwidth value of all the allocated bandwidth, and the packet traffic on the equal bandwidth can only be forwarded to the associated business processing instance to avoid impact Traffic from other transponders.
  • the step 103 includes:
  • Step 201 Receive a message sent by a transponder, determine the mapping identifier carried in the message according to the mapping strategy and the encapsulation type, query the scheduling unit lookup table according to the mapping identifier, determine the corresponding scheduling unit, and pass the scheduling The unit schedules the message.
  • Step 202 When the scheduling result of the scheduling unit is passed, query the forwarding table of the business processing instance through the mapping identifier, determine the corresponding business processing instance, and forward the message to the business processing instance.
  • Step 203 When the scheduling result of the scheduling unit is discarded, discard the packet, and count the discarded packet statistically.
  • FIG. 5 it is a schematic diagram of a message forwarding process carrying a mapping identifier.
  • the packet carrying the mapping identifier is forwarded to the LB of the control plane, first query the mapping strategy and encapsulation type, then parse out the mapping identifier from the packet according to the mapping strategy and encapsulation type, and use the mapping identifier as a key to the scheduling unit lookup table
  • the scheduling unit is found in, and the message is input to the inquired scheduling unit to schedule the message. There are two scheduling results: pass or discard.
  • mapping identifier is used as a keyword to query the forwarding table of the business processing instance, the destination business processing instance is obtained, and finally the message is forwarded to the destination business processing instance for control business processing. When it is discarded, the discarded packets are counted.
  • the cluster management component of the control plane obtains performance data of LB and business processing instances.
  • the performance data of LB includes: the total bandwidth of each scheduling unit of the scheduling queue, the currently occupied bandwidth, and the number of packet losses;
  • the performance data of the service processing instance includes: the total bandwidth of the interface, the currently occupied bandwidth, the number of packet losses, CPU utilization, Memory usage.
  • the total bandwidth on the LB generally does not change.
  • the cluster management component can calculate the total bandwidth of each scheduling unit, and the subsequent cluster management component does not need to obtain the LB scheduling unit.
  • the total bandwidth The total bandwidth of the interface of the business processing instance generally does not change.
  • the cluster management component is notified of the total bandwidth of the interface, and the subsequent cluster management component does not need to obtain the total bandwidth of the interface of the business processing instance.
  • Other performance data is monitored in real time by the cluster management component.
  • the monitoring methods include pull mode and push mode: the pull mode is that the cluster management component periodically queries the performance data of LB and each business processing instance. In this way, the cluster management component can control the frequency of acquiring performance data, and the cluster management component will not work overloaded.
  • the real-time performance is poor; the push mode is when the LB or business processing instance perceives the performance data changes, and actively informs the cluster management component of the changed performance data.
  • This method has high real-time performance, but when the performance data changes drastically, the cluster management There is a lot of data sent by the component, and there is a risk of overloading the cluster management component.
  • the cluster management component When the cluster management component detects a change in the performance of the LB or business processing instance, the packet loss count of the scheduling unit for LB increases, the packet loss count for the interface of the business processing instance, the CPU utilization increase exceeds the defined threshold, and the memory usage rate exceeds the definition Threshold, the cluster management component dynamically adjusts the forwarding of LB forwarded packets to mitigate and solve performance problems.
  • the method further includes:
  • Step 104 Adjust at least one of the equal bandwidth corresponding to the repeater and the number of service processing instances in real time according to the load.
  • the embodiment of the present disclosure adjusts the occupied bandwidth of the transponder according to the load condition through an automatic monitoring and adjustment mechanism, thereby improving bandwidth utilization and system throughput.
  • the first case the load situation includes the packet loss of the scheduling unit, and the equal bandwidth corresponding to the repeater is adjusted.
  • the bandwidth to be increased is determined according to the number of lost packets and the length of the packet loss; the weight of the equal bandwidth corresponding to the transponder is adjusted according to the bandwidth to be increased, or, Increase the equal bandwidth corresponding to the repeater.
  • the dynamic adjustment may include the following steps:
  • Step 301 after the LB initialization is completed, the cluster management component is notified of the total bandwidth of the LB.
  • the cluster management component divides the bandwidth of the LB into multiple equal bandwidths according to the granularity defined by the user policy, and allocates a scheduling unit for each equal bandwidth.
  • the bandwidth set by the scheduling unit is the bandwidth value of the corresponding equal bandwidth.
  • Step 302 when the cluster management component detects that the LB equal bandwidth scheduling unit has packet loss, the LB actively reports or the cluster management component queries the performance data of the current LB packet loss equal bandwidth scheduling unit. Trigger the dynamic adjustment processing of the forwarded message forwarding.
  • the cluster management component calculates the bandwidth to be increased on the repeater based on the packet loss count on the scheduling unit and the time factor. There are two ways to increase the newly added bandwidth: migrate the newly added bandwidth to other scheduling units of the repeater (adjust the weight of the equal bandwidth corresponding to the repeater according to the increased bandwidth); repeater allocation New equal bandwidth, which increases the total bandwidth for forwarding packets sent by the transponder on the LB to avoid packet loss of packets sent by the transponder.
  • the cluster management component determines that it is a condition for migrating to another scheduling unit and adding bandwidth to another scheduling unit: when the repeater has other allocated equal bandwidth, and the remaining bandwidth of the other allocated equal bandwidth can meet the newly added bandwidth bandwidth.
  • the forwarding message of the repeater is adjusted to select the equal bandwidth scheduling unit.
  • step 304 the cluster management component notifies the repeater management component of the equal-weight bandwidth ID of the changed weight and the new weight value.
  • Step 305 The repeater management component receives the equal bandwidth ID and the weight value, queries the mapping identifier according to the equal bandwidth ID, and sets a new selection weight value for the mapping identifier, that is, the weight value of the equal bandwidth. Update the forwarding packet encapsulation table of the transponder, set the new selection weight of the mapping identifier, and notify the forwarder to update the forwarding packet encapsulation table.
  • Step 306 when the repeater to which the scheduling unit where the packet is lost has only one equal bandwidth, or the remaining bandwidth of other equal bandwidths on the repeater does not meet the bandwidth to be added, and the total equal bandwidth of the repeater does not reach the allocatable bandwidth If the upper limit is exceeded, equal bandwidth processing is added for the repeater.
  • the cluster management component allocates equal bandwidth from the total bandwidth of LB, and obtains the equal bandwidth ID and equal bandwidth weight. If the total equal bandwidth of the transponder reaches the allocatable upper limit and the adjustment cannot be continued, an alarm can be given to exit the adjustment process.
  • the cluster management component obtains the equal bandwidth ID and the equal bandwidth weight to notify the forwarder management component.
  • the repeater management component generates a mapping identifier for the equal bandwidth according to the mapping strategy and the encapsulation strategy, and specifies the weight of the equal bandwidth as the selection weight of the mapping identifier.
  • a new packet encapsulation table entry is generated, and the mapping identifier and selection weight are added to the entry.
  • Step 308 the forwarder is notified of the newly added packet encapsulation table entry, and the forwarder receives and adds it to the local packet encapsulation table.
  • Step 309 the repeater management component notifies the cluster management component of the mapping identifier.
  • the cluster management component adds a new scheduling unit look-up table entry for the newly added equal bandwidth, and fills in the mapping identifier and scheduling unit of the equal bandwidth.
  • the cluster management component allocates business processing instances for the newly added equal bandwidth: when the business processing instance associated with the transponder to which the newly added equal bandwidth has spare processing capacity, the business processing instance is selected as the newly added equal bandwidth Business processing instance; otherwise, a new business processing instance is allocated for the newly added equal bandwidth.
  • the cluster management component generates a business processing instance forwarding table entry, adds a mapping identifier for the newly added equal bandwidth, and selects or adds a new business processing instance for the newly added equal bandwidth.
  • Step 310 The repeater management component notifies the LB of the newly generated scheduling unit lookup table entry, and the LB receives the scheduling unit lookup table entry and adds it to the local scheduling unit lookup entry.
  • step 311 the repeater management component notifies the LB of the newly generated service processing instance forwarding table entry, and the LB receives the control instance forwarding table entry and adds it to the local service processing instance forwarding table.
  • the load condition includes the performance condition of the service processing instance, adjusting at least one of the equal bandwidth corresponding to the repeater, and the number of service processing instances.
  • the performance problem may include, but is not limited to: packet loss on the interface, CPU utilization exceeding a preset threshold, and memory utilization exceeding a preset threshold.
  • the dynamic adjustment may include the following steps:
  • Step 401 after each business processing instance in the cluster is initialized, the business processing instance notifies the cluster management component of the total bandwidth of the interface of the business processing instance, and the cluster management component receives the total bandwidth of the interface of each business processing instance to save.
  • Step 402 a performance problem occurs in the service processing instance: packet loss occurs on the interface, the CPU utilization exceeds a defined threshold, and the memory utilization exceeds a preset threshold.
  • the business processing instance actively informs the cluster management component or the cluster management component to periodically query the performance data of the business processing instance, and the cluster management component triggers the adjustment processing of the business processing instance.
  • the cluster management component calculates the capacity of the newly added business processing instance, and checks whether the remaining capabilities of other business processing instances on the transponder of the newly added business processing instance can meet the capacity of the newly added business processing instance.
  • the cluster management component determines that the remaining capacity of other business processing instances of the transponder meets the capacity of the newly added business processing instance, and adjusts the weight of the equal bandwidth associated with the business processing instance of the transponder, which can increase the service in addition to performance problems.
  • the weight of the equal bandwidth associated with other business processing instances of the processing instance can also reduce the weight of the equal bandwidth associated with the business processing instance that has performance problems.
  • Steps 404 to 405 are the same as steps 304 to 305 and will not be repeated here.
  • step 406 the cluster management component determines that the remaining capacity of other business processing instances of the transponder does not meet the capacity of the newly added business processing instance.
  • the cluster management component allocates a new equal bandwidth and business processing instance from the total bandwidth of the LB and obtains the allocation The weight of equal bandwidth.
  • Steps 407 to 408 are the same as steps 307 to 308 and will not be repeated here.
  • the repeater management component notifies the cluster management component of the mapping identifier.
  • the cluster management component adds a new scheduling unit look-up table entry for the newly added equal bandwidth, and fills in the mapping identifier and scheduling unit of the equal bandwidth.
  • the cluster management component generates a business processing instance forwarding table entry, adds a mapping identifier for the newly added equal bandwidth, and adds a new business processing instance for the newly added equal bandwidth.
  • Steps 410 to 411 are the same as steps 310 to 311 and will not be repeated here.
  • the load condition includes the bandwidth occupied by the scheduling unit, and the equal portion of bandwidth allocated to the repeater is recovered.
  • the bandwidth occupied by each scheduling unit is obtained; when the scheduling unit associated with the repeater is greater than one, it is determined whether to reclaim the equal bandwidth associated with the repeater according to the bandwidth occupied by the scheduling unit associated with the repeater.
  • the cluster management component monitors each equal-part bandwidth scheduling unit associated with the repeater and obtains the bandwidth occupied by each scheduling unit. If there is only one scheduling unit associated with the repeater, equal bandwidth recovery is not performed. When it is greater than one, first obtain the scheduling unit with the smallest occupied bandwidth from all scheduling units, and then calculate the remaining bandwidth sum of other remaining scheduling units. If the remaining bandwidth sum is greater than or equal to the occupied bandwidth of the scheduling unit, an equal bandwidth recovery process is performed .
  • the sum of the remaining bandwidths of the scheduling units other than the scheduling unit with the smallest occupied bandwidth is greater than the occupied bandwidth of the scheduling unit with the smallest occupied bandwidth, and the recovery is determined The equal bandwidth associated with the scheduling unit with the smallest occupied bandwidth.
  • the equal bandwidth associated with the transponder After the equal bandwidth associated with the transponder is recovered, check whether all the equal bandwidth remaining after the recovery of the transponder needs to be recovered, and if the remaining equal bandwidth meets the recovery conditions, continue to recover until the transponder has only one equal portion remaining
  • the bandwidth or the occupied bandwidth of each scheduling unit of the repeater is greater than the sum of the remaining bandwidths of other scheduling units.
  • the recovering the equal bandwidth associated with the repeater may include:
  • Identifying that the equal bandwidth to be recovered is idle recovering the mapping identifier of the idle equal bandwidth, deleting an entry packet encapsulation table entry related to the idle equal bandwidth, and notifying the forwarder to delete the idle bandwidth Encapsulation table entry related to the equal bandwidth transmission of the equal part, and delete the idle lookup table entry related to the equal bandwidth related scheduling unit entry and the service processing instance forwarding table entry.
  • the cluster management component identifies that the portion of the bandwidth is idle, and notifies the forwarder management component to reclaim the mapping identifier of the equal bandwidth. After retrieving the mapping identifier, delete the relevant encapsulation table entry of the protocol packet and notify the forwarder to delete the transmission. Protocol packet encapsulation table entry.
  • the cluster management component deletes the scheduling unit query table entry and the service processing instance forwarding table entry associated with the equal bandwidth, and notifies the LB to delete the scheduling unit query table entry and the business processing instance forwarding table entry.
  • the bandwidth on the LB of each repeater is adjusted in real time according to the load, so that a certain repeater can not consume the LB bandwidth and achieve the LB Full use of bandwidth.
  • the bandwidth on the LB of each repeater is adjusted in real time according to the load, so that a certain repeater can not consume the LB bandwidth and achieve the LB Full use of bandwidth.
  • by monitoring the load of business processing instances in real time and dynamically adjusting the workload it avoids overloading of a specific business processing instance.
  • Step 501 a cluster discovery phase, which is the first phase of startup. After the one or more transponders managed by the control plane are started, the transponder management component is notified of joining, and the transponder management component receives the joining notification of each transponder and registers the transponder information. Inside the control plane, the repeater component, LB, and one or more business processing instances notify the cluster management component that the cluster management component receives the component join notification and registers the component information.
  • Step 502 in the configuration and initialization phase, the system is initialized after configuration and configuration.
  • the configuration is mainly: granularity of equal bandwidth, mapping strategy and encapsulation strategy of equal bandwidth, threshold of CPU usage of business processing instance, and threshold of memory usage of business processing instance.
  • the equalizer bandwidth is allocated to the forwarder, and the LB forwarding and forwarding packet entry is generated according to the equal bandwidth, which may include: mapping The strategy and encapsulation strategy table, the scheduling unit lookup table, and the service processing instance forwarding table.
  • the forwarder sends the packet to the LB by querying these entries for forwarding.
  • a forwarding packet encapsulation table corresponding to the equal bandwidth when the forwarding packet of the forwarder is forwarded to the LB is generated.
  • the forwarding packet selects the appropriate mapping identifier according to the selection weight of the mapping identifier, and transmits the packet according to the encapsulation type
  • the encapsulation map identifier is forwarded to LB.
  • the performance data of the LB and the performance data of the service processing instance are monitored in real time.
  • the forwarded packet of the transponder is forwarded and the service processing is adjusted to solve the performance problem. It can include: adjusting the selection weight of the mapping identifier on the transponder corresponding to the equal bandwidth, adjusting the flow; or adding an equal bandwidth and business processing instance to the transponder to adjust the bandwidth and business processing capacity allocated to the transponder to solve the performance problem.
  • the automatic adjustment phase is an iterative phase. The optimal performance optimization that one adjustment may not be possible, and then recursively adjust multiple times according to the performance data until the performance problem is solved optimally. Artificial intelligence (AI) technology can also be added during the automatic adjustment stage to optimize performance intelligently, automatically, and flexibly.
  • AI Artificial intelligence
  • the load sharing method of the embodiment of the present disclosure which is applied to the cluster management component, includes:
  • Step 601 The cluster management component divides the total bandwidth of the packets carried by the load balancer LB into multiple equal bandwidths.
  • the cluster management component divides the total bandwidth of the LB message into multiple equal bandwidths according to a preset granularity.
  • the cluster management component divides the total bandwidth of the LB into multiple equal bandwidths according to the granularity defined by the user policy, and assigns an ID to each equal bandwidth.
  • the ID of each equal bandwidth is unique and cannot be different from other The equal bandwidth is the same.
  • step 601 the method further includes:
  • the cluster management component instructs the LB to generate a scheduling unit according to the equal bandwidth, each equal portion of the bandwidth corresponds to a scheduling unit, so that the LB uses the scheduling unit to limit the bandwidth, and the scheduling unit limits the bandwidth
  • the value is the bandwidth value of the corresponding equal bandwidth.
  • the scheduling unit is located in the LB, and specific implementation includes, but is not limited to: software implementation of the general server, programmable hardware implementation, and hardware circuit implementation. These implementations are specifically related to business scenarios, performance, and cost considerations.
  • Step 602 The cluster management component allocates one or more equal bandwidths to the repeater, the one or more equal bandwidths are associated with the repeater, and the allocated one Or multiple equal bandwidths are associated with the service processing instance, so that the packet sent by the repeater to the LB is forwarded to the corresponding service processing instance through the equal bandwidth associated with the repeater.
  • the cluster management component allocates one or more equal bandwidths to the repeater, including:
  • the cluster management component determines the minimum number of equal bandwidths required by the repeater according to the bandwidth requirements of the repeater, and allocates the minimum number of equal bandwidths to the repeater.
  • the cluster management component calculates the minimum number of equal bandwidths, the maximum number of equal bandwidths, and the default equal bandwidth weights required by the repeaters according to the bandwidth requirements of the forwarders sent by the forwarder management components, and then assigns the minimum number of equal portions bandwidth.
  • the cluster management component allocates one or more equal bandwidths to the repeater, further including:
  • the cluster management component allocates the equal bandwidth identifier and the equal bandwidth weight to the one or more equal bandwidths, and sends the equal bandwidth identifier and the equal bandwidth weight to the repeater management component to Causing the transponder management component to generate an encapsulation table of the transmitted message.
  • the upload packet encapsulation table specifies a field carried by the upload packet.
  • the field identifies an equal bandwidth.
  • the equal bandwidth ID is used to generate the field according to a mapping strategy.
  • the field is a mapping identifier.
  • the method further includes:
  • the cluster management component generates a scheduling unit query entry based on the correspondence between the equal bandwidth associated with the repeater and the mapping identifier, and the correspondence between the equal bandwidth and the scheduling unit, and the scheduling unit query entry constitutes the scheduling unit Lookup table.
  • the cluster management component queries the scheduling unit according to the equal bandwidth ID, generates a scheduling unit query entry, the key of the entry is the mapping identifier, and the query result is the scheduling unit, and notifies the LB of the scheduling unit query entry.
  • the associating the allocated one or more equal portions of bandwidth with the service processing instance includes:
  • the cluster management component generates a business processing instance forwarding table according to the correspondence between the equal bandwidth associated with the repeater and the mapping identifier, and the correspondence between the mapping identifier and the business processing instance.
  • the cluster management component generates a business processing instance forwarding table
  • the key of the entry is a mapping identifier
  • the query result is a business processing instance.
  • the processing capability of a business processing instance is relatively strong, so initially, multiple transponders can correspond to one business processing instance. That is to say, there is a case where different mapping identifiers query the same business processing instance for an entry in the business processing instance forwarding table.
  • the method further includes:
  • the cluster management component adjusts the equal bandwidth corresponding to the repeater in real time according to the load of the LB.
  • the load condition of the LB includes a packet loss condition of the scheduling unit of the LB
  • the cluster management component adjusts the equal bandwidth corresponding to the repeater in real time according to the load condition of the LB, including:
  • the cluster management component determines the bandwidth to be increased according to the number of lost packets and the length of the lost packets;
  • the cluster management component adjusts the weight of the equal bandwidth corresponding to the repeater according to the increased bandwidth, or increases the equal bandwidth corresponding to the repeater.
  • the load situation of the LB includes the bandwidth situation of the scheduling unit of the LB, and the cluster management component adjusts the equal bandwidth corresponding to the repeater in real time according to the load situation of the LB, including:
  • the cluster management component obtains the bandwidth occupied by each scheduling unit
  • the cluster management component determines whether to reclaim the equal bandwidth associated with the repeater according to the bandwidth occupied by the scheduling unit associated with the repeater.
  • the cluster management component determines, based on the bandwidth occupied by the scheduling unit associated with the forwarder, whether to reclaim the equal bandwidth associated with the forwarder, including:
  • the cluster management component determines that the sum of the remaining bandwidths of the scheduling units other than the scheduling unit with the smallest occupied bandwidth among all scheduling units associated with the repeater is greater than the occupied bandwidth of the scheduling unit with the smallest occupied bandwidth Retrieve the equal bandwidth associated with the scheduling unit with the smallest occupied bandwidth.
  • the method further includes:
  • the cluster management component adjusts in real time at least one of the equal bandwidth corresponding to the repeater and the number of business processing instances according to the load of the business processing instances.
  • the load situation of the business processing instance includes the performance situation of the business processing instance
  • the cluster management component adjusts the equal bandwidth and service of the forwarder in real time according to the load situation of the business processing instance
  • At least one of the number of processing instances includes:
  • the cluster management component determines that the capacity of the business processing instance needs to be increased when the performance problem of the business processing instance occurs
  • the cluster management component performs at least one of the following operations according to the increased business processing instance capability required:
  • the cluster management component manages the bandwidth reported by the LB, and also manages the performance of business processing instances. As shown in FIG. 10, the cluster management component may include a component management module 701 and a forwarding management module 702.
  • the component management module 701 can be set as:
  • the components include LB, all business processing instances and repeater management components.
  • the LB performance data includes: the total bandwidth of the LB connection repeater; the total bandwidth of all scheduling units of the scheduling queue, the used bandwidth, and the number of lost packets.
  • the performance data of the service processing instance includes: interface bandwidth, used bandwidth and packet loss of the interface; CPU usage rate and memory usage rate.
  • the performance data is sent to the forwarding management module 702 of the cluster management component.
  • the forwarding management module 702 can be set to:
  • the configuration may include: an equal bandwidth mapping strategy and a encapsulation type; a business processing instance's CPU usage threshold exceeding the threshold; a business processing instance's memory usage threshold exceeding the threshold; an equal bandwidth bandwidth granularity. Notify the LB user of the configured mapping strategy and encapsulation type.
  • each requested repeater is allocated the minimum number of equal bandwidths and the maximum number of equal bandwidths. Returns the minimum number of equal bandwidths of the repeater requested by the repeater management component, and the ID of each equal bandwidth and the initial weight of equal bandwidth.
  • the weight of the equal bandwidth of the repeater is notified to the repeater management component. If the calculated remaining bandwidth does not meet the newly added bandwidth, the repeater is allocated a new equal bandwidth, the repeater management component is notified of the newly allocated equal bandwidth ID and weight, and the equal return returned by the repeater management component is received Bandwidth mapping identification. Generate a newly allocated equal-part bandwidth scheduling unit look-up table entry and add the returned mapping identifier and equal-part bandwidth scheduling unit. Allocate business processing instances for the newly-added equal bandwidth.
  • the business processing instance allocated to the repeater will be adopted, otherwise, new business processing will be selected For instance, generate a forwarding table entry for the business processing instance, fill in the mapping identifier of the equal bandwidth and select the business processing instance.
  • Performance problems trigger workload adjustment processing for business processing instances. Calculating business performance instances that have performance problems based on performance data requires new capabilities. Check whether the remaining capabilities of other business processing instances of the transponder to which the business processing instance belongs can meet the need to add new capacity, adjust the equal bandwidth weights associated with the business processing instances of the transponder, and reduce the business processing instances with performance problems The weight of the associated equal bandwidth, or increase the weight of the associated equal bandwidth of other business processing instances, and notify the forwarder management component of the changed equal bandwidth weight.
  • a new equal bandwidth and a new business processing instance are allocated, the repeater management component is notified of the equal bandwidth ID and weight, and the repeater management component is received Mapping ID generated based on equal bandwidth ID and weight, mapping strategy, encapsulation strategy.
  • a new scheduling unit look-up table entry is generated, the mapping unit corresponding to the allocated equal bandwidth and the equal bandwidth scheduling unit are filled in, and the generated entry is notified to the LB.
  • Generate a new business processing instance forwarding table entry fill in the mapping identifier corresponding to the allocated equal bandwidth and a new business processing instance, and notify the LB of the generated entry.
  • the load sharing method of the embodiment of the present disclosure which is applied to LB, includes:
  • Step 801 the LB receives the message sent by the repeater; wherein, the repeater is associated with one or more equal bandwidths, and the one or more equal bandwidths are associated with the service processing instance.
  • the method before step 801, the method further includes:
  • Step 800 The LB generates a scheduling unit according to the equal bandwidth according to the instruction of the cluster management component, and each equal bandwidth corresponds to a scheduling unit; the LB uses the scheduling unit to limit the bandwidth and the scheduling unit limits the speed
  • the bandwidth value of is the bandwidth value of the corresponding equal bandwidth.
  • the scheduling unit is located in the LB, and specific implementation includes, but is not limited to: software implementation of the general server, programmable hardware implementation, and hardware circuit implementation. These implementations are specifically related to business scenarios, performance, and cost considerations.
  • Step 802 the LB forwards the message to the corresponding service processing instance through the equal bandwidth associated with the forwarder.
  • step 802 includes:
  • Step 901 The LB determines the mapping identifier carried in the message according to the mapping strategy and the encapsulation type, queries the scheduling unit lookup table according to the mapping identifier, determines the corresponding scheduling unit, and reports the report through the scheduling unit Text for scheduling.
  • Step 902 When the scheduling result of the scheduling unit is passed, the LB queries the forwarding table of the business processing instance through the mapping identifier, determines the corresponding business processing instance, and forwards the message to the business processing instance.
  • Step 903 When the scheduling result of the scheduling unit is discarded, the LB discards the packet, and counts the discarded packet.
  • the LB also notifies the cluster management component of real-time performance data, where the performance data includes: the occupied bandwidth of the scheduling unit, the packet loss count of the scheduling unit, and the time factor of the packet loss count.
  • the LB receives a notification from the cluster management component, and adds or deletes a scheduling unit lookup table entry.
  • the LB receives a notification from the cluster management component, and adds or deletes a business processing instance forwarding table entry.
  • the LB implemented in the present disclosure includes a forwarding module 1001 and a resource management module 1002.
  • the forwarding module 1001 may be set as:
  • the resource management module 1002 may be set as:
  • the performance data includes: the occupied bandwidth of the scheduling unit, the packet loss count of the scheduling unit, and the time factor of the packet loss count, etc.
  • the scheduling unit query table entry addition or deletion request received by the cluster management component is notified, and the LB forwarding module is notified to add or delete the scheduling unit query table entry.
  • the load sharing method implemented in the present disclosure which is applied to the repeater management component, includes:
  • step 1101 the repeater management component determines one or more equal bandwidths allocated by the cluster management component to the repeater.
  • the method before step 1101, the method further includes: the repeater management component determines a bandwidth requirement of the repeater, and notifies the cluster management component of the bandwidth requirement.
  • the repeater management component analyzes and estimates the repeater bandwidth requirement of the repeater, and notifies the cluster management component of the repeater bandwidth requirement.
  • the repeater management component receives the notification of the cluster management component and learns that the cluster management component allocates one or more equal bandwidth IDs and equal bandwidth weights for the repeater.
  • Step 1102 The repeater management component associates the allocated one or more equal bandwidths with the repeater, so that the packet sent by the repeater to the LB is related to the repeater.
  • the connected equal bandwidth is forwarded to the corresponding business processing instance.
  • the repeater management component associates the allocated one or more equal portions of bandwidth with the repeater, including:
  • the repeater management component receives the cluster management component to allocate the equal bandwidth identifiers and equal bandwidth weights to the one or more equal bandwidths, and generates an upload based on the equal bandwidth identifiers and equal bandwidth weights A packet encapsulation table, sending the packet encapsulation table to the forwarder.
  • the generating of the encapsulation table of the transmitted packet according to the identifier of the equal bandwidth and the weight of the equal bandwidth includes:
  • the repeater management component determines a mapping identifier according to the identifier of the equal bandwidth, and the encapsulation table of the uploaded packet includes the mapping identifier and the equal bandwidth bandwidth weight.
  • the method further includes:
  • the repeater management component updates the upload packet encapsulation table according to the notification of the cluster management component, and notifies the repeater to update the upload packet encapsulation table.
  • the repeater management component of the control plane discovers the repeater and manages the performance reported by the repeater.
  • the repeater management component of the embodiment of the present disclosure includes: a resource management module 1201 and a forwarding management module 1202.
  • the resource management module 1201 may be set as:
  • the forwarding management module 1202 may be set to:
  • a business processing example of an embodiment of the present disclosure includes: a business processing module 1301 and a resource management module 1302.
  • the business processing module 1301 may be set to:
  • the resource management module 1302 may be set as:
  • the load sharing method of the embodiment of the present disclosure which is applied to the repeater, includes:
  • step 1401 the repeater obtains the encapsulation table of the uploaded packet; wherein, the encapsulation table of the transmitted packet indicates that the repeater is associated with one or more equal bandwidths.
  • Step 1402 the repeater sends the message to the LB according to the encapsulation table of the uploaded message, so that the message is forwarded to the corresponding service process through the equal bandwidth associated with the repeater Examples.
  • the encapsulation table of the uploaded message includes a mapping identifier and an equal bandwidth weight.
  • the forwarder sends the message to the LB according to the encapsulated table of the uploaded message, including:
  • the repeater selects a corresponding equal bandwidth for the packet according to the equal bandwidth weight, determines a corresponding mapping identifier according to the selected equal bandwidth, and encapsulates the mapping identifier according to a preset packaging strategy In the message, it is sent to the LB.
  • the method further includes:
  • the repeater updates the encapsulation table of the transmitted message according to the notification of the repeater management component.
  • the repeater includes: a forwarding module 1501 and a resource management module 1502.
  • the forwarding module 1501 may be set as:
  • the resource management module 1502 may be set as:
  • the total forwarder message flow can be controlled to avoid affecting the flow of other forwarders, and through the automatic monitoring adjustment mechanism,
  • the occupied bandwidth of the transponder is adjusted according to the system load, which improves bandwidth utilization and system throughput.
  • FIG. 20 it is a networking diagram of an implementation mode: multiple PPPoE (Point-to-Point Protocol Over Ethernet, Ethernet-based point-to-point communication protocol) Client (client) access to the operator's access
  • the access network is connected to the transponder upward.
  • the transponder is located at the edge layer of the metropolitan area network, and the control plane of the vBRAS is located at the data center in the metropolitan area network.
  • the PPPoE client sends the PPPoE access message to the vBRAS control plane through the repeater to complete the PPPoE client access control.
  • the control plane is composed of two business processing instances, and the PPPoE client's protocol transmission message is processed in the business processing instance.
  • the granularity of the set equal bandwidth is to process 10 upload packets per second, dividing 30 equal bandwidths, which corresponds to the scheduling unit that generates 30 equal bandwidths.
  • the cluster management component allocates a minimum number of equal bandwidths of 1, a maximum number of equal bandwidths of 10, and a weight of equal bandwidth of 10.
  • the mapping strategy of the repeater management component is to map the equal bandwidth ID to a VLAN, the mapping identifier of the first equal bandwidth allocated by the repeater is VLAN 10, and the encapsulation strategy is to insert it behind the VLAN field of the original packet.
  • the bandwidth-forwarded packets are forwarded to service processing instance 1 for processing.
  • FIG. 21 is a schematic flowchart of forwarding a forwarded message to a service processing unit after allocating an equal portion of bandwidth.
  • step 1601 the PPPoE client sends a PPPoE protocol message, and the repeater receives the PPPoE protocol message.
  • Step 1602 Query the forwarding packet encapsulation table on the forwarder. Because the forwarder only allocates one equal bandwidth, the mapping ID obtained is VLAN 10, and the encapsulation strategy is to insert it into the message. If a VLAN exists, insert VLAN 10 after the existing VLAN.
  • Step 1603 the PPPoE packet encapsulating VLAN 10 is forwarded to the LB of the control plane.
  • Step 1604 parse out the equal bandwidth mapping identifier VLAN 10 from the packet, use VLAN 10 to query the scheduling unit lookup table, obtain the scheduling unit to schedule the packet, and the current scheduling unit concurrent protocol report The number of documents is less than 10, so the remaining bandwidth can meet the message, then forward the message, use VLAN 10 to query the query table of the service processing instance, and obtain service processing instance 1.
  • Step 1605 the PPPoE protocol packet is forwarded to the service processing instance 1, and the PPPoE protocol packet is processed on the service processing instance 1.
  • the scheduling unit When the bandwidth of the scheduling unit of equal bandwidth cannot meet the received message, the scheduling unit discards the packet and counts the packet loss.
  • the cluster management component detects and adjusts the packet loss by the scheduling unit, and solves the problem. The forwarding performance of the text in LB.
  • 22 is a flowchart of automatically adjusting the bandwidth of the transponder when a performance problem is detected.
  • step 1701 the PPPoE client sends a PPPoE protocol message, and the repeater receives the PPPoE protocol message.
  • Step 1702 the repeater inserts the equal bandwidth mapping ID VLAN 10 into the PPPoE protocol packet, and then forwards the packet to the LB.
  • Step 1703 the LB parses out the mapping identifier VLAN 10 from the packet according to the mapping strategy and the encapsulation strategy, and uses VLAN 10 to query the scheduling unit.
  • the scheduling unit is queried to schedule PPPoE packets. Since the scheduling unit has now sent 10 protocols concurrently Packets, so the bandwidth usage has reached the upper limit, so the protocol packets cannot be forwarded, the packets are discarded, and the packet loss count is counted.
  • step 1704 the cluster management component monitors that the LB scheduling unit has lost packets, and triggers automatic adjustment of service packet forwarding.
  • step 1705 the cluster management component allocates a new equal bandwidth to the repeater, and then allocates a scheduling unit of the equal bandwidth.
  • the total bandwidth of the scheduling unit is the bandwidth value of the equal bandwidth.
  • step 1706 the cluster management component notifies the repeater management component of the ID and weight of the equal bandwidth.
  • Step 1707 the repeater management component generates a mapping identifier VLAN 20 according to the mapping strategy for the new equal bandwidth of the repeater.
  • the repeater management component In step 1708, the repeater management component generates an entry of the forwarding packet encapsulation table on the forwarder, adds the newly generated mapping identifier VLAN 20, weight and encapsulation strategy, and notifies the forwarder to add the entry of the forwarding packet encapsulation table generated.
  • step 1709 the repeater management component notifies the cluster management component of the generated mapping identifier VLAN 20.
  • Step 1710 The cluster management component generates a new equal-bandwidth scheduling unit look-up table entry, adds a scheduling unit mapping mapping VLAN 20 and equal-bandwidth, and notifies the LB to add the entry, and the scheduling unit look-up table entry query keyword To map VLAN 20, the query result is the scheduling unit.
  • Service processing instance 1 has remaining capacity.
  • the newly created equal-bandwidth business processing instance selects business processing instance 1, generates a new equal-bandwidth business processing instance forwarding table entry, adds a mapping identifier VLAN 20 and business processing instance 1, And notify the LB to add the entry, the query keyword of the business processing instance forwarding table entry is the mapping identification VLAN 20, and the query result is business processing instance 1.
  • Step 1711 The PPPoE client protocol packet is discarded before the adjustment, the PPPoE client retransmits the PPPoE protocol packet, and the retransmitted packet is forwarded to the repeater.
  • Step 1712 the forwarding packet encapsulation table of the repeater has two entries, and the mapping identifiers are VLAN 10 and VLAN 20 respectively. According to the selection strategy of the encapsulation table on the forwarder, VLAN 20 is selected for encapsulation of the mapping identifier. The packet is forwarded to LB.
  • the LB parses the mapping identifier VLAN 20 from the packet, and queries the scheduling unit through the VLAN 20. Since the equal bandwidth of the scheduling unit is newly added, the scheduling unit has sufficient bandwidth to forward the protocol packet. Then, query the forwarding table of the business processing instance according to the mapping identifier VLAN 20.
  • the destination business processing instance is business processing instance 1, and forward PPPoE protocol packets to the business processing instance 1 for processing.
  • FIG. 23 is a flowchart of an example of automatically adjusting business processing when packet loss occurs on the interface of the business processing instance.
  • step 1801 the PPPoE client sends a PPPoE protocol message, and the repeater receives the PPPoE protocol message.
  • Step 1802 the repeater inserts the equal bandwidth mapping ID VLAN 20 into the PPPoE protocol packet, and then forwards the packet to the LB.
  • Step 1803 the LB parses the mapping identifier VLAN 20 from the packet according to the mapping strategy and the encapsulation strategy, queries the scheduling unit with the VLAN 20, and finds that the scheduling unit schedules PPPoE packets, and the available bandwidth of the scheduling unit satisfies PPPoE protocol packets. And then forwarded to business processing instance 1.
  • Step 1804 the interface bandwidth of service processing instance 1 is used up, PPPoE protocol packets are discarded, and the packet loss count is counted.
  • step 1805 the cluster management component monitors that packet loss occurs on the interface of the business processing instance 1, and triggers a business processing instance adjustment process.
  • step 1806 packet loss occurs on the interface of the business processing instance 1, and the business processing instance 2 is added to share the workload of the business processing instance 1. Because there are two equal bandwidths of the associated business instance 1, the equal bandwidth with the mapping ID of VLAN 20 is selected Related business processing example 2.
  • the cluster management component uses VLAN 20 as a keyword to query the forwarding table of the business processing instance to obtain forwarding entries, update the business processing instance to business processing instance 2, and notify the LB to update the forwarding table entry of the business processing instance.
  • step 1808 the PPPoE client protocol packet is discarded before the adjustment, the PPPoE client retransmits the PPPoE protocol packet, and the retransmitted packet is forwarded to the repeater.
  • step 1809 the repeater encapsulates the mapping identifier VLAN 20 in the PPPoE protocol packet and forwards it to the LB.
  • Step 1810 LB parses out the mapping identifier VLAN 20 from the protocol packet, and finds the scheduling unit corresponding to the equal bandwidth.
  • the scheduling unit has available bandwidth to forward the protocol packet.
  • the protocol packet searches the service processing instance forwarding table through VLAN 20 to obtain
  • the target business processing instance is business processing instance 2, and PPPoE protocol packets are sent to business processing instance 2 for processing.
  • FIG. 24 is a schematic flowchart of the replacement of the package type.
  • Step 1901 the DHCP client applies for an address and sends out a DHCP protocol message.
  • Step 1902 the repeater receives the DHCP protocol message. Query the encapsulation table of sent packets, and query the equal bandwidth mapping identifier and encapsulation type.
  • the mapping identifier is encapsulated VLAN200, and the encapsulation type is replacement.
  • the original packet carries VLAN100. Because the encapsulation type is replacement, VLAN100 is stripped first and then VLAN200 is encapsulated.
  • Step 1903 the repeater sends a DHCP protocol message to the LB, carrying VLAN 200.
  • Step 1904 on the LB, the VLAN 200 is parsed according to the mapping strategy and the encapsulation type, and the VLAN 200 is used as the mapping identifier to query the scheduling unit lookup table. After the scheduling unit schedules the packet to be forwarded, the business processing instance lookup table is queried to obtain the business processing instance 1.
  • Step 1905 LB sends a DHCP protocol message to service processing instance 1.
  • FIG. 25 is a schematic flow chart of the package type inserted into the tail.
  • Step 2001 The DHCP client applies for an address and sends out a DHCP protocol message.
  • Step 2002 the repeater receives the DHCP protocol message. Query the encapsulation table of sent packets, and query the equal bandwidth mapping identifier and encapsulation type.
  • the mapping identifier is encapsulated VLAN200, and the encapsulation type is inserted into the tail.
  • the original message carries VLAN100. Since the encapsulation type is inserted into the tail, VLAN200 is encapsulated after VLAN100.
  • step 2003 the repeater sends a DHCP protocol message to the LB, carrying VLAN 100 and VLAN 200.
  • Step 2004, on the LB locate the VLAN 200 whose mapping identifier is the VLAN 100 of the packet according to the mapping strategy and encapsulation type, and parse out the VLAN 200.
  • VLAN 200 is used as the mapping identifier to query the scheduling unit lookup table. After the scheduling unit schedules the packet, the packet is forwarded and queried Business processing instance query table to obtain business processing instance 1.
  • Step 2005 LB sends a DHCP protocol message to service processing instance 1.
  • the protocol packet carries a mapping identifier and is mapped to a VLAN, and it can also specify fields mapped to other packets. For example, a dedicated line user goes online through ARP (Address Resolution Protocol), and protocol packets are transmitted between the repeater and the LB through the VxLAN tunnel, and the mapping identifier can be mapped to the VNI of VxLAN.
  • ARP Address Resolution Protocol
  • Fig. 26 is a schematic flowchart.
  • Step 2101 the dedicated line user sends an ARP protocol message.
  • Step 2102 the repeater receives the ARP protocol message. Query the encapsulation table of sent packets, and query the equal bandwidth mapping identifier and encapsulation type.
  • the mapping identifier is package VNI10. After the protocol packet is encapsulated by VxLAN, set the VNI to 10.
  • Step 2103 the repeater sends an ARP protocol packet to the LB.
  • the ARP protocol packet has been VxLAN encapsulated, and the VNI is 10.
  • Step 2104 the LB receives the ARP protocol packet encapsulated by VxLAN, decapsulates the VxLAN, obtains VNI 10, determines VNI 10 as the mapping identifier according to the mapping strategy and encapsulation type, and queries the scheduling unit lookup table according to VNI 10, after the scheduling unit After scheduling, the message is forwarded, query the query table of the business processing instance, and obtain the business processing instance 1.
  • Step 2105 LB sends an ARP protocol packet to service processing instance 1.
  • transponders send packets to the same control plane. If the transponder traffic is heavy, it will occupy more bandwidth on the control plane entrance LB, which will affect the packets sent by other transponders. Setting the maximum number of equal bandwidths for each transponder can limit the maximum flow of packets sent by each transponder, so as to ensure that it does not affect other transponders. As shown in FIG. 27, there are 20 equal bandwidths on the LB, the repeater 1 allocates equal bandwidths of 1 to 10 equal bandwidths, and the repeater 2 allocates equal bandwidths of 1 to 10 equal bandwidths.
  • Figure 28 is a flow chart of the traffic limit of the packets sent by the repeater within the range.
  • step 2201 an equal portion of bandwidth is allocated to repeater 1 and repeater 2 in the initialization phase.
  • Step 2202 the forwarded packet traffic of forwarder 1 is forwarded to the LB, and the forwarded traffic does not exceed the allocated equal bandwidth.
  • Step 2203 the LB forwards the packet traffic of the forwarder 1 to the service processing instance.
  • Step 2204 the forwarded packet traffic of forwarder 1 is forwarded to the LB, and the forwarded traffic exceeds the allocated equal bandwidth.
  • step 2205 the traffic of the forwarding packet of the forwarder 1 is scheduled on the LB. Since the traffic exceeds the allocated equal bandwidth, the excess is discarded, and the packet loss count is counted.
  • step 2206 the cluster management component monitors and finds that there is a packet loss in the packet sent by the LB forwarder 1 and allocates a new equal bandwidth.
  • Step 2207 the forwarded packet traffic of the forwarder 1 is forwarded to the LB. Since the forwarder 1 allocates a new equal bandwidth, the forward traffic does not exceed the allocated equal bandwidth.
  • Step 2208 the LB forwards the packet traffic of the forwarder 1 to the service processing instance.
  • Step 2209 the forwarding packet flow of the forwarder 1 is scheduled on the LB, because the flow exceeds the allocated equal bandwidth, the excess is discarded, and the packet loss count is counted
  • step 2210 since 10 equal bandwidths have been allocated to the repeater 1, the equal bandwidth cannot be allocated again when the upper limit of allocation is reached.
  • step 2211 the message sent by the repeater 2 is forwarded to the LB.
  • the repeater 2 has allocated an equal portion of bandwidth, which can meet the scheduled forwarding of the transmitted message, and the transmitted message is forwarded to the service processing instance.
  • an embodiment of the present disclosure further provides a control plane entity, including: a memory 2301, a processor 2302, and a computer program 2303 stored on the memory 2301 and executable on the processor 2302, and the processor 2302
  • the method of load sharing shown in FIG. 3 is implemented when the program is executed.
  • an embodiment of the present disclosure further provides a cluster management component, including: a memory 2401, a processor 2402, and a computer program 2403 stored on the memory 2401 and executable on the processor 2402, and the processor 2402
  • the method of load sharing shown in FIG. 9 is implemented when the program is executed.
  • an embodiment of the present disclosure also provides an LB, including: a memory 2501, a processor 2502, and a computer program 2503 stored on the memory 2501 and executable on the processor 2502.
  • the processor 2502 executes When the program is described, the method of load sharing shown in FIG. 11 is realized.
  • an embodiment of the present disclosure also provides a repeater management component, including: a memory 2601, a processor 2602, and a computer program 2603 stored on the memory 2601 and executable on the processor 2602, the processor 2602 A method for implementing load sharing as shown in FIG. 15 when the program is executed.
  • an embodiment of the present disclosure further provides a transponder, including: a memory 2701, a processor 2702, and a computer program 2703 stored on the memory 2701 and executable on the processor 2702, the processor 2702 executes The program implements the load sharing method shown in FIG. 18.
  • Embodiments of the present disclosure also provide a computer-readable storage medium that stores computer-executable instructions that are used to perform the load sharing method.
  • the above storage medium may include, but is not limited to: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc.
  • ROM read-only memory
  • RAM random access memory
  • mobile hard disk magnetic disk or optical disk, etc.
  • computer storage media includes both volatile and nonvolatile implemented in any method or technology for storing information such as computer readable instructions, data structures, program modules, or other data Sexual, removable and non-removable media.
  • Computer storage media include but are not limited to RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, magnetic tape, magnetic disk storage or other magnetic storage devices, or may Any other medium for storing desired information and accessible by a computer.
  • the communication medium generally contains computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium .
  • the load sharing method, control plane entity and repeater provided by the embodiments of the present invention have the following beneficial effects:
  • the packet traffic on the repeater reaches the control plane, it can only occupy the bandwidth value of all the allocated bandwidth
  • the packet traffic on the equal bandwidth can only be forwarded to the associated business processing instance to avoid affecting the traffic of other transponders.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

Selon des modes de réalisation, la présente invention concerne un procédé de partage de charge, une entité de plan de commande et un répéteur. Le procédé comprend les étapes suivantes : un plan de commande divise la bande passante totale pour prendre en charge des paquets en de multiples bandes passantes égales ; le plan de commande attribue une ou plusieurs bandes passantes égales à un répéteur, associe la ou les bandes passantes égales attribuées au répéteur et associe la ou les bandes passantes égales attribuées à une instance de traitement de service ; le plan de commande reçoit un paquet envoyé par le répéteur et le transmet à l'instance de traitement de service correspondante au moyen de la ou des bandes passantes égales associées au répéteur. Dans les modes de réalisation de la présente invention, lorsque le flux de paquets sur un répéteur atteint un plan de commande, seules les valeurs de bande passante de toutes les bandes passantes égales attribuées peuvent être occupées et le flux de paquets sur les bandes passantes égales ne peut être transmis qu'à une instance de traitement de service associée, évitant ainsi d'influencer le flux d'autres répéteurs.
PCT/CN2019/124319 2018-12-10 2019-12-10 Procédé de partage de charge, entité de plan de commande et répéteur WO2020119682A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811505039.3A CN109787911B (zh) 2018-12-10 2018-12-10 负载分担的方法、控制面实体和转发器
CN201811505039.3 2018-12-10

Publications (1)

Publication Number Publication Date
WO2020119682A1 true WO2020119682A1 (fr) 2020-06-18

Family

ID=66496837

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/124319 WO2020119682A1 (fr) 2018-12-10 2019-12-10 Procédé de partage de charge, entité de plan de commande et répéteur

Country Status (2)

Country Link
CN (1) CN109787911B (fr)
WO (1) WO2020119682A1 (fr)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109787911B (zh) * 2018-12-10 2021-05-04 中兴通讯股份有限公司 负载分担的方法、控制面实体和转发器
CN115696455A (zh) * 2021-07-27 2023-02-03 华为技术有限公司 一种流控处理的方法及通信装置
CN116700956B (zh) * 2023-05-23 2024-02-23 海易科技(北京)有限公司 请求处理方法、装置、电子设备和计算机可读介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130301460A1 (en) * 2006-06-30 2013-11-14 Centurylink Intellectual Property Llc System and method for adjusting codec speed in a transmission path during call set-up due to reduced transmission performance
CN105610632A (zh) * 2016-02-14 2016-05-25 华为技术有限公司 一种虚拟网络设备及相关方法
CN107666419A (zh) * 2016-07-28 2018-02-06 中兴通讯股份有限公司 一种虚拟宽带接入方法、控制器和系统
CN107786613A (zh) * 2016-08-30 2018-03-09 新华三技术有限公司 宽带远程接入服务器bras转发实现方法和装置
CN109787911A (zh) * 2018-12-10 2019-05-21 中兴通讯股份有限公司 负载分担的方法、控制面实体和转发器

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105100142B (zh) * 2014-05-04 2018-10-09 新华三技术有限公司 软件定义网络协议报文的传输控制方法及装置
US9369200B1 (en) * 2014-12-18 2016-06-14 Juniper Networks, Inc. Network controller having predictable analytics and failure avoidance in packet-optical networks
CN106055380B (zh) * 2016-05-20 2019-04-26 郑州丞极信息科技有限责任公司 一种业务服务器的整合方法及系统
CN106941456B (zh) * 2017-05-17 2019-08-30 华中科技大学 一种软件定义网络中控制平面的负载均衡方法及系统
CN107979539B (zh) * 2017-10-30 2021-12-03 新华三技术有限公司 报文处理方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130301460A1 (en) * 2006-06-30 2013-11-14 Centurylink Intellectual Property Llc System and method for adjusting codec speed in a transmission path during call set-up due to reduced transmission performance
CN105610632A (zh) * 2016-02-14 2016-05-25 华为技术有限公司 一种虚拟网络设备及相关方法
CN107666419A (zh) * 2016-07-28 2018-02-06 中兴通讯股份有限公司 一种虚拟宽带接入方法、控制器和系统
CN107786613A (zh) * 2016-08-30 2018-03-09 新华三技术有限公司 宽带远程接入服务器bras转发实现方法和装置
CN109787911A (zh) * 2018-12-10 2019-05-21 中兴通讯股份有限公司 负载分担的方法、控制面实体和转发器

Also Published As

Publication number Publication date
CN109787911A (zh) 2019-05-21
CN109787911B (zh) 2021-05-04

Similar Documents

Publication Publication Date Title
US11895154B2 (en) Method and system for virtual machine aware policy management
US11381507B2 (en) Virtual network device and related method
US10938748B2 (en) Packet processing method, computing device, and packet processing apparatus
US8949413B2 (en) Filter selection and resuse
WO2020119682A1 (fr) Procédé de partage de charge, entité de plan de commande et répéteur
US10122623B2 (en) Control device and control method in SDN network
CN107306232B (zh) 网络设备、控制器、队列管理方法及流量管理芯片
JP5677588B2 (ja) マルチチャネルパケット伝送のためのシステムおよび方法
US11438236B2 (en) Methods and systems for managing connected data transfer sessions
US20220286409A1 (en) Method and apparatus for configuring quality of service policy for service, and computing device
US20160021005A1 (en) Communication system, control apparatus and communication apparatus
US20200359255A1 (en) Quality of service (qos) control method and related device
US20170310493A1 (en) Network entity and service policy management method
WO2020083272A1 (fr) Procédé et système de génération de stratégie de traitement, et support d'informations
KR20170033179A (ko) 소프트웨어 정의 네트워크 기반 가상 네트워크 사용 대역폭 관리 방법 및 가상 네트워크 관리장치
WO2016150020A1 (fr) Procédé et dispositif de programmation de paquets à base d'identificateur de flux de programmation
US9749240B2 (en) Communication system, virtual machine server, virtual network management apparatus, network control method, and program
CN108683598B (zh) 一种非对称网络流量处理方法和处理装置
US10015133B2 (en) Value-added service deployment method, network device, and value-added service device
US11799676B2 (en) Deterministic assignment of overlay multicast traffic
CN114710975A (zh) 多域间传输多传输网络上下文标识
CN110868357A (zh) 网络流量控制方法、vtep设备和存储介质
WO2022237860A1 (fr) Procédé de traitement de paquets, procédé d'attribution de ressources et dispositif associé
CN114915590A (zh) 一种报文转发的方法、装置和存储介质
WO2016023149A1 (fr) Procédé permettant de garantir une qualité de service d'un trafic et contrôleur central

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19894735

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 22/10/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19894735

Country of ref document: EP

Kind code of ref document: A1