US20060087968A1 - Load balancing in data networks - Google Patents

Load balancing in data networks Download PDF

Info

Publication number
US20060087968A1
US20060087968A1 US10/519,241 US51924105A US2006087968A1 US 20060087968 A1 US20060087968 A1 US 20060087968A1 US 51924105 A US51924105 A US 51924105A US 2006087968 A1 US2006087968 A1 US 2006087968A1
Authority
US
United States
Prior art keywords
hash
routing
parameters
logic
packets
Prior art date
Legal status (The legal status 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 status listed.)
Abandoned
Application number
US10/519,241
Other languages
English (en)
Inventor
Hoyt Bowen
Patrick Droz
Lukas Kencl
Bernhard Metzler
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
International Business Machines Corp
Original Assignee
International Business Machines Corp
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 International Business Machines Corp filed Critical International Business Machines Corp
Assigned to INTERNATIONAL BUSINESS MACHINES CORPORATION reassignment INTERNATIONAL BUSINESS MACHINES CORPORATION ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: BOWEN, HOYT EDWIN JR., DROZ, PATRICK, METZLER, BERNARD, KENCL, LUKAS
Publication of US20060087968A1 publication Critical patent/US20060087968A1/en
Priority to US12/118,830 priority Critical patent/US7769025B2/en
Abandoned legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/22—Alternate routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
    • H04L9/40—Network security protocols
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H04L45/123—Evaluation of link metrics
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/24—Multipath
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/38—Flow based routing
    • 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
    • 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
    • H04L47/11—Identifying congestion
    • 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
    • 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
    • H04L47/17—Interaction among intermediate nodes, e.g. hop by hop
    • 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
    • H04L47/30—Flow control; Congestion control in combination with information about buffer occupancy at either end or at transit nodes
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1012—Server selection for load balancing based on compliance of requirements or conditions with available server resources
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1023—Server selection for load balancing based on a hash applied to IP addresses or costs
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1029—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers using data related to the state of servers by a load balancer
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/10015—Access to distributed or replicated servers, e.g. using brokers

Definitions

  • the present invention generally relates to load balancing in data networks and particularly relates to a method and apparatus for load balancing in a data network.
  • a data network typically comprises a plurality of server computers interconnected to a plurality of client computers via a data communications network infrastructure.
  • the network infrastructure typically comprises a plurality of intermediate data communications devices or nodes such as switches, routers and the like for routing data packets between the client computer and the server computers.
  • Such data communications devices typically comprise a plurality of input/output (I/O) ports, a switch fabric for routing data packets received on one port to one or more of the other ports, and control logic for controlling the switch fabric to make appropriate connections between the ports based on address information contained in the transient data packets.
  • I/O input/output
  • a problem associated with such data networks is that of balancing loads between different nodes within the network infrastructure. As the amount of data traffic handled by the network infrastructure increases, the balance between loads carried by the nodes in the infrastructure becomes harder to maintain in a timely manner. Accordingly, communication bottlenecks and delays are incurred.
  • Load balancing apparatus for a data communications network, the apparatus comprising: hash logic for computing a hash function on incoming data packets; a threshold detector connected to the hash logic for triggering, in response to utilization of the downstream objects exceeding a predefined threshold, redefinition in the hash logic of parameters of the hash function from a first set of parameters to a second set of parameters for redistributing the data packets amongst the downstream objects; wherein, the hash logic, in use, directs the packets for routing to downstream objects in the network via a first routing path based on a hash computation using the first set of parameters, and, if the threshold is exceeded, for selectively directing the packets to one of the first routing path and a second routing path in dependence on separate hash computations using the first and the second sets of parameters for subsequent routing of the packets via the selected one of the first and second routing paths based on the results of one of the separate hash computations.
  • the hash logic in use directs the data packet to the first routing path if the results of the separate hash computations intersect and otherwise directs the data packet to the second routing path.
  • the apparatus further comprises a filter connected to the hash logic for selectively bypassing the hash logic for flows having a lifetime exceeding a predefined value.
  • the apparatus further comprises the first routing path and the second routing path, the first routing path comprising first routing logic connected to the hash logic, and the second routing path comprising second routing logic connected to the hash logic, wherein the first routing path is faster than the second routing path, and wherein, on the second routing path, downstream objects are selected based on packet flow status.
  • the first routing logic may comprise at least one network processor and the second routing logic may comprise at least one general purpose processor.
  • the second routing logic may be configured to detect a flow delimiter in a flow of data packets and, on detection of the start indictor, to route the corresponding flow according to the hash computation using the second parameters.
  • the second routing logic may also be configured to detect flows of packets exceeding a predetermined inactivity time and to route such flows according to the hash computation using the second parameters. Further, the second routing logic may be configured to detect flows of packets exceeding a predetermined lifetime and to direct such flows to the first routing logic.
  • the present invention extends to an application specific integrated circuit comprising load balancing apparatus as herein before described.
  • the present invention also extends to a network infrastructure node comprising load balancing apparatus as herein before described.
  • the present invention extends to a data communications network comprising such a network infrastructure node.
  • a method of load balancing in a data communications network comprising: computing a hash function on incoming data packets; triggering, in response to utilization of the downstream objects exceeding a predefined threshold, redefinition of parameters of the hash function from a first set of parameters to a second set of parameters for redistributing the data packets amongst the downstream objects; and, directing the packets for routing to downstream objects in the network via a first routing path based on a hash computation using the first set of parameters, and, if the threshold is exceeded, selectively directing the packets to one of the first routing path and a second routing path in dependence on separate hash computations using the first and the second sets of parameters for subsequent routing of the packets via the selected one of the first and second routing paths based on the results of one of the separate hash computations.
  • the method comprises directing the data packets to the first routing path if the separate hash computations intersect and otherwise directing the data packets to the second routing path.
  • an iterative hardware load balancing technique in which a hash function is employed to balance packet flows between a number of hosts, intermediate nodes, and/or network interfaces in a data communications network.
  • the hash is performed on a portion of each packet that is constant for the duration of a flow, such as the address of the source associated with the flow.
  • the technique alternates between a state in which the hash function is computed with only one set of hash parameters and a state in which two hash parameter sets are given. In one state, one hash result is known. In the other state, the difference between two hash results is known.
  • the technique identifies a routing fast path and a routing slow path.
  • the routing fast path may be performed in dedicated hardware such as Network Processors or similar application specific integrated circuits (ASICs).
  • the routing slow path may be better performed in software on a general purpose processor (GPP).
  • GPS general purpose processor
  • No data flows in progress are moved between load balanced objects unless continuity is maintained. This advantageously insures that flow connectivity is uninterrupted and that packets are not reordered.
  • state is retained for data flows outside the intersection of the two hash functions. This kept retained state is however constantly and aggressively reduced. This advantageously minimizes hardware costs.
  • by performing the fast path routing via dedicated hardware such as network processor hardware, data communications performance is improved beyond hitherto available levels.
  • FIG. 1 is a block diagram of a data communications network
  • FIG. 2 is a block diagram of an infrastructure node of the data communications network
  • FIG. 3 is another block diagram of the infrastructure node
  • FIG. 4 is a flow chart associated with a control point of the infrastructure node
  • FIG. 5 is a flow chart associated with a load balancer of the infrastructure node.
  • FIG. 6 is yet another flow chart associated with a load balancer of the infrastructure node.
  • a data communications network 140 comprises a plurality of network objects 10 - 120 .
  • the network objects include network nodes 10 - 90 and intervening network links 100 - 120 for communicating data packets between the network nodes 10 - 30 .
  • the network nodes 10 - 90 include end point nodes 10 - 60 and infrastructure nodes 70 - 90 .
  • the infrastructure nodes 70 - 90 of the links 110 - 130 collectively form a network infrastructure 130 .
  • the end point nodes 10 - 60 include a plurality of client data processing devices 10 - 30 and a plurality of server computer systems 40 - 60 .
  • the clients 10 - 30 and the servers 40 - 60 are interconnected via an intervening network infrastructure 100 .
  • Each client 10 - 30 may be a personal computer, laptop computer, personal digital assistant, mobile telephone, or the like.
  • Each server 40 - 60 may be a file server, print server, or similar host data processing or data storage resource.
  • data communications between the clients 10 - 30 and the servers 40 - 60 is effected via flows of data packets through the network infrastructure 100 .
  • the flows of data packets are passed between the infrastructure nodes 70 - 90 of the network infrastructure en route between the clients 10 - 30 and the servers 40 - 60 according to one or more data communications protocol.
  • Each infrastructure node 70 - 90 performs a packet routing function to forward data packets it receives to an appropriate recipient object
  • each infrastructure node 70 - 90 comprises first routing logic 220 , second routing logic 210 , and a load balancer 200 connected to both the first routing logic 220 and the second routing logic 210 .
  • Both the first routing logic 220 and the second routing logic 210 perform the same routing function. However, the first routing logic 220 performs the routing function faster than the second routing logic 210 .
  • the first routing logic 220 thus provides a relatively fast routing path and the second routing logic 210 provides a relatively slow routing path.
  • each infrastructure node comprises a network processor (NP) 300 connected to a control point 400 .
  • NP network processor
  • Both the load balancer 200 and the first routing logic 220 are implemented in the NP 300 .
  • the NP 300 comprises executable software for performing load balancing decisions of the load balancer 200 , together with a hardwired logic implementation of first routing logic 220 for performing the packet routing function.
  • the load balancer 200 comprises a filter 230 and hash logic 240 connected to the filter 230 .
  • a control point (CP) 400 connected to the NP 300 comprises a general purpose processing unit (GPP) 410 connected to a memory 420 .
  • the memory 420 is stored computer program code for implementing, on execution by the GPP 410 , a flow state controller 430 , threshold detectors 440 and 490 , a hash parameter generator 450 , a conflicting flow state table 460 , a ‘long living’ flows state table 480 , and a router 470 .
  • the flow state controller 430 computes state for client-server flows.
  • the threshold detector 440 monitors the current load status of the servers 40 - 60 to provide load management.
  • the router 470 implements the second routing logic 210 . Performance of the packet routing function by the CP 400 is slower than performance of the packet routing function by the NP 300 .
  • the load balancer 200 determines whether each flow of incoming data packets is processed by the first routing logic 220 or by the CP 400 in dependence on current status of the filter 230 and the hash logic 240 .
  • the status of both the filter 230 and the hash logic 240 is controlled by the CP 400 based on prevailing demands on one or more of the downstream objects in the network 140 .
  • Such load balancing is desirable in the interests of, for example, preventing unnecessary overloading of one or more of the downstream objects by non optimally distributed client requests. By preventing unnecessary overloading, the available service level in the network 140 is optimized.
  • the load balancer 200 via the hash logic 240 therein, performs hash functions to balance flows of data packets between objects in the network.
  • the hash logic 240 performs the hash functions on some portion of the incoming data packets that is constant for the duration of a packet flow, such as the source address.
  • the load balancer 200 alternates between a state in which the difference between the results of two hash functions is known, and a state in which the result of only one hash function is known.
  • the filter 230 may bypass the hash logic 240 for some packet flows.
  • the filter 230 directs incoming flows of data packets either to the first routing logic 220 or the hash logic 240 .
  • the hash logic 240 in dependence on a control input from the CP 400 , directs the packets either to the first routing logic 220 or to the CP 400 in which the second routing logic 210 is implemented. If the packet is passed to the first routing logic 220 , then it is coupled with routing information from either the hash logic 240 or the filter 230 . The routing information added directs the first routing logic 220 to choose a specified routing path. If the hash function redirects a packet to the second routing logic 210 , then the routing information is also included in the packet.
  • the control point 400 starts in step 500 with the initialization of both the hash logic 240 and the filter 230 of the load balancer 200 .
  • the filter 230 is set to void.
  • the initial parameters of the hash logic 240 may be based on operator configuration. Alternatively, the initial hash parameters may be based on an automated query of resource capacities of objects 40 - 60 to be balanced. Examples of such resource capabilities include CPU speeds, interface speeds, and the like.
  • the CP 400 accumulates statistics in the feedback memory 430 relating to resource utilization in the objects 40 - 60 to be balanced.
  • the threshold detector 440 checks the statistics accumulated in the flow state controller 430 against a predetermined threshold to determine if at least one of the objects 40 - 60 is over utilized relative to the other objects. In the event that the determination is positive, then, at step 530 , new hash parameters are computed by the hash parameter generator 450 and loaded as a second parameter set into the hash logic 240 . The new hash parameters optimally distribute the network traffic based on the statistics gathered.
  • the CP 400 acts as the second routing logic 210 . Specifically, the CP 400 receives packets from the load balancer 200 which produced different hash results when the hash logic 240 computed the hash function with both the old and the new parameter set.
  • the CP 400 establishes and maintains per flow state information within the conflicting flow state table 460 . If the CP 400 does not receive packets for a conflicting flow within a configurable time interval, or if the CP 400 at least once receives a flow end indicator such as a TCP FIN bit, then the flow in question is regarded as closed and its state is marked as ‘old terminated’.
  • a flow end indicator such as a TCP FIN bit
  • step 545 the CP 400 forwards the packet through the second routing logic 470 to one of the load balanced objects 40 - 60 . If the status of the flow to which the packet belongs is marked as ‘old terminated’, then the packet is forwarded to the result of the hash computation using the new hash parameters. Otherwise, the packet is forwarded to the result of the hash computation using the old hash parameters.
  • the threshold detector 490 checks, at step 550 , the number of unterminated conflicting flows against a predetermined threshold. If the threshold detector 490 detects a smaller or equal amount of flows than set as the threshold, then at step 560 all local state of unterminated conflicting flows is moved to the ‘long living’ flows state table 480 and the conflicting flow state table 460 is cleared. In this case, at step 560 , the contents of the ‘long living’ flows state table 480 is transferred to the filter function 230 of the load balancer 200 and the old hash parameters of the hash logic 240 are removed. If, at step 550 , the threshold detector 490 detects more unterminated conflicting flows than set as the threshold, then the CP 400 keeps receiving hash conflicting packets from the load balancer 200 .
  • the CP 400 receives copies of packets belonging to ‘long living’ flows.
  • the CP 400 checks to determine if a packet carries a flow end indicator such as a TCP FIN bit or if a ‘long living’ flow was inactive for a predetermined interval. If both determinations are negative, then the CP 400 returns to step 570 , continuing to receive copies of packets belonging to ‘long living’ flows.
  • the CP 400 checks if the ‘long living’ flow table 480 is empty. If so, the status of the CP 400 returns to the accumulation of feedback information in step 510 . If the ‘long living’ flow table 480 is not empty, then the CP 400 returns to step 570 , receiving copies of packets belonging to ‘long living’ flows.
  • Deep packet processing may be employed by the CP 400 to identify flows in progress which are to be considered as terminated. For example, in some circumstances, File Transfer Protocol (FTP) flows in progress may be routed to a new server without harm.
  • FTP File Transfer Protocol
  • the filter 230 checks if a filter rule is set for each incoming packet. If yes, then, in step 610 , the filter 230 performs a match of a packet flow identifier against the local filter rules. If no filter rule exists or no rule matches, then, in step 620 the packet is forwarded to the hash logic 240 . If the packet matches a filter rule, then, based on the matching filter rule, in step 630 the packet is coupled with its routing information. In step 640 the packet is copied to the CP 400 and in step 650 the packet is sent to the first routing logic 220 .
  • the hash logic 240 alternates between two states, where either one set of hashing parameters is known or one old set and one new set of hashing parameters are known.
  • the hash logic 240 computes the hash result based on the new set of parameters.
  • the hash logic 240 checks to determine if a set of old hash parameters exists. If no old parameter set exists, then in step 720 the packet and the hash result are passed to the first routing logic 220 to send the packet to the appropriate target object 40 - 60 . If an old set of hash parameters exists, then in step 730 , the hash result is recomputed based on the old parameter set.
  • step 740 the hash result is checked to determine if both hash computations yield the same result. If so, then the packet is passed in step 720 to the first routing logic 220 to be routed to the appropriate target object 40 - 60 . If not, then in step 750 , the packet and both hash results are passed to the CP 400 for slow path routing.
  • each infrastructure 90 - 110 in the network 130 comprises load balancing functionality.
  • networks 130 may be assembled in which only a subset of infrastructure nodes comprise load balancing functionality.
  • Load balancing apparatus for a data communications network comprises hash logic for computing a hash function on incoming data packets.
  • a threshold detector is connected to the hash logic for triggering, in response to utilization of the downstream objects exceeding a predefined threshold, redefinition in the hash logic of parameters of the hash function from a first set of parameters to a second set of parameters for redistributing the data packets amongst the downstream objects.
  • the hash logic in use, directs the packets for routing to downstream objects in the network via a first routing path based on a hash computation using the first set of parameters, and, if the threshold is exceeded, for selectively directing the packets to one of the first routing path and a second routing path in dependence on separate hash computations using the first and the second sets of parameters for subsequent routing of the packets via the selected one of the first and second routing paths based on the results of one of the separate hash computations.
  • load balancing apparatus 200 for a data communications network comprises hash logic 240 for computing a hash function on incoming data packets.
  • a threshold detector 440 is connected to the hash logic 240 for triggering, in response to utilization of the downstream objects exceeding a predefined threshold, redefinition in the hash logic 240 of parameters of the hash function from a first set of parameters to a second set of parameters for redistributing the data packets amongst the downstream objects.
  • the hash logic 240 directs the packets for routing to downstream objects in the network via a first routing path based on a hash computation using the first set of parameters, and, if the threshold is exceeded, selectively directs the packets to one of the first routing path and a second routing path in dependence on separate hash computations using the first and the second sets of parameters for subsequent routing of the packets via the selected one of the first and second routing paths based on the results of one of the separate hash computations. It will be appreciated however, that the present invention is not limited to two sets of hash parameters. More than two sets of hash parameters may be employed by the hash logic 240 in other embodiments of the present invention.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Hardware Design (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
US10/519,241 2002-06-24 2003-06-20 Load balancing in data networks Abandoned US20060087968A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US12/118,830 US7769025B2 (en) 2002-06-24 2008-05-12 Load balancing in data networks

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
EP02014116.4 2002-06-24
EP02014116 2002-06-24
PCT/IB2003/002409 WO2004002109A1 (fr) 2002-06-24 2003-06-20 Equilibrage des charges dans des reseaux de donnees

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US12/118,830 Continuation US7769025B2 (en) 2002-06-24 2008-05-12 Load balancing in data networks

Publications (1)

Publication Number Publication Date
US20060087968A1 true US20060087968A1 (en) 2006-04-27

Family

ID=29797140

Family Applications (2)

Application Number Title Priority Date Filing Date
US10/519,241 Abandoned US20060087968A1 (en) 2002-06-24 2003-06-20 Load balancing in data networks
US12/118,830 Expired - Fee Related US7769025B2 (en) 2002-06-24 2008-05-12 Load balancing in data networks

Family Applications After (1)

Application Number Title Priority Date Filing Date
US12/118,830 Expired - Fee Related US7769025B2 (en) 2002-06-24 2008-05-12 Load balancing in data networks

Country Status (7)

Country Link
US (2) US20060087968A1 (fr)
EP (1) EP1516475B1 (fr)
JP (1) JP4452176B2 (fr)
CN (1) CN100539580C (fr)
AU (1) AU2003232413A1 (fr)
DE (1) DE60303384T2 (fr)
WO (1) WO2004002109A1 (fr)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060233106A1 (en) * 2005-04-14 2006-10-19 Microsoft Corporation Stateless, affinity-preserving load balancing
US20080240400A1 (en) * 2005-09-02 2008-10-02 Data Connection Limited Telephone call processing method and apparatus
US20120271964A1 (en) * 2011-04-20 2012-10-25 Blue Coat Systems, Inc. Load Balancing for Network Devices
US20140337417A1 (en) * 2011-12-19 2014-11-13 Solbox Inc Method and apparatus for processing server load balancing by using the result of hash function
US9225651B2 (en) 2011-07-08 2015-12-29 Telefonaktiebolaget L M Ericsson (Publ) Method and apparatus for load balancing
CN106656842A (zh) * 2015-10-29 2017-05-10 华为数字技术（苏州）有限公司 一种负载均衡方法和流量转发设备

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7764612B2 (en) * 2005-06-16 2010-07-27 Acme Packet, Inc. Controlling access to a host processor in a session border controller
CN100407705C (zh) * 2006-04-12 2008-07-30 华为技术有限公司 一种路由器控制方法和系统
US7684352B2 (en) * 2006-11-02 2010-03-23 Nortel Networks Ltd Distributed storage of routing information in a link state protocol controlled network
US8218553B2 (en) * 2009-02-25 2012-07-10 Juniper Networks, Inc. Load balancing network traffic on a label switched path using resource reservation protocol with traffic engineering
US8259585B1 (en) * 2009-04-17 2012-09-04 Juniper Networks, Inc. Dynamic link load balancing
US20100274893A1 (en) * 2009-04-27 2010-10-28 Sonus Networks, Inc. Methods and apparatus for detecting and limiting focused server overload in a network
US8699343B2 (en) 2009-04-27 2014-04-15 Sonus Networks, Inc. Adaptive rate control based on overload signals
US9923798B1 (en) 2012-06-28 2018-03-20 Juniper Networks, Inc. Dynamic load balancing of network traffic on a multi-path label switched path using resource reservation protocol with traffic engineering
US9331891B2 (en) * 2012-10-11 2016-05-03 International Business Machines Corporation Virtual consolidated appliance
US10230621B2 (en) 2017-05-09 2019-03-12 Juniper Networks, Inc. Varying a per-hop-bandwidth constraint in multi-path label switched paths

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6044080A (en) * 1996-11-19 2000-03-28 Pluris, Inc. Scalable parallel packet router
JP3369445B2 (ja) * 1997-09-22 2003-01-20 富士通株式会社 ネットワークサービスサーバ負荷調整装置、方法および記録媒体
GB2349296B (en) * 1999-04-21 2001-04-04 3Com Corp Reduction of imbalance in transmit traffic queues in a network switch
CA2328011A1 (fr) * 2000-01-18 2001-07-18 Lucent Technologies Inc. Methode et appareil pour equilibrer la charge des services de reseau
JP4150159B2 (ja) * 2000-03-01 2008-09-17 富士通株式会社 伝送経路制御装置及び伝送経路制御方法並びに伝送経路制御プログラムを記録した媒体
US6553005B1 (en) * 2000-07-26 2003-04-22 Pluris, Inc. Method and apparatus for load apportionment among physical interfaces in data routers

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060233106A1 (en) * 2005-04-14 2006-10-19 Microsoft Corporation Stateless, affinity-preserving load balancing
US7693050B2 (en) * 2005-04-14 2010-04-06 Microsoft Corporation Stateless, affinity-preserving load balancing
US8134916B2 (en) 2005-04-14 2012-03-13 Microsoft Corporation Stateless, affinity-preserving load balancing
US20080240400A1 (en) * 2005-09-02 2008-10-02 Data Connection Limited Telephone call processing method and apparatus
US8611522B2 (en) * 2005-09-02 2013-12-17 Metaswitch Networks Ltd Telephone call processing method and apparatus
US20120271964A1 (en) * 2011-04-20 2012-10-25 Blue Coat Systems, Inc. Load Balancing for Network Devices
US9705977B2 (en) * 2011-04-20 2017-07-11 Symantec Corporation Load balancing for network devices
US9225651B2 (en) 2011-07-08 2015-12-29 Telefonaktiebolaget L M Ericsson (Publ) Method and apparatus for load balancing
US20140337417A1 (en) * 2011-12-19 2014-11-13 Solbox Inc Method and apparatus for processing server load balancing by using the result of hash function
US9525730B2 (en) * 2011-12-19 2016-12-20 Solbox Inc Method and apparatus for processing server load balancing by using the result of hash function
CN106656842A (zh) * 2015-10-29 2017-05-10 华为数字技术（苏州）有限公司 一种负载均衡方法和流量转发设备

Also Published As

Publication number Publication date
CN1659845A (zh) 2005-08-24
WO2004002109A8 (fr) 2004-03-18
EP1516475B1 (fr) 2006-01-25
JP4452176B2 (ja) 2010-04-21
US20080219258A1 (en) 2008-09-11
DE60303384D1 (de) 2006-04-13
DE60303384T2 (de) 2006-08-17
WO2004002109A1 (fr) 2003-12-31
CN100539580C (zh) 2009-09-09
JP2005536087A (ja) 2005-11-24
EP1516475A1 (fr) 2005-03-23
US7769025B2 (en) 2010-08-03
AU2003232413A1 (en) 2004-01-06

Similar Documents

Publication Publication Date Title
US7769025B2 (en) Load balancing in data networks
US10938748B2 (en) Packet processing method, computing device, and packet processing apparatus
KR102536676B1 (ko) 패킷 처리 방법 및 장치, 및 관련 디바이스들
US8325607B2 (en) Rate controlling of packets destined for the route processor
US9001827B2 (en) Methods for configuring network switches
US20170048148A1 (en) Method, apparatus, and system for load balancing of service chain
CN106713182B (zh) 一种处理流表的方法及装置
US8787388B1 (en) System and methods for forwarding packets through a network
US20160261507A1 (en) Method and apparatus for controlling and managing flow
US11489815B2 (en) Methods and systems for synchronizing state amongst monitoring nodes
CN112311674B (zh) 报文发送方法、装置及存储介质
CN106656857A (zh) 一种报文限速的方法和装置
US6973503B2 (en) Preventing at least in part control processors from being overloaded
US10944695B2 (en) Uplink port oversubscription determination
JP2022166934A (ja) 情報処理装置、過負荷制御プログラムおよび過負荷制御方法
CN111769994A (zh) 一种基于软件定义网络的网络功能快速扩容方法
CN107113244B (zh) 一种数据转发的方法、装置和系统
Mon et al. Flow path computing in software defined networking
CN111404705A (zh) 一种sdn的优化方法、装置及计算机可读存储介质
EP4440064A1 (fr) Procédé et appareil de communication, et système de communication
CN118540273A (zh) 转发报文方法和装置
CN118138542A (zh) 一种数据处理的方法和相关设备

Legal Events

Date Code Title Description
AS Assignment

Owner name: INTERNATIONAL BUSINESS MACHINES CORPORATION, NEW Y

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:BOWEN, HOYT EDWIN JR.;DROZ, PATRICK;KENCL, LUKAS;AND OTHERS;REEL/FRAME:016984/0528;SIGNING DATES FROM 20050118 TO 20050215

STCB Information on status: application discontinuation

Free format text: ABANDONMENT FOR FAILURE TO CORRECT DRAWINGS/OATH/NONPUB REQUEST