CN102355422B - Multicore, parallel and lock-free quality of service (QOS) flow control method - Google Patents
Multicore, parallel and lock-free quality of service (QOS) flow control method Download PDFInfo
- Publication number
- CN102355422B CN102355422B CN201110315458.2A CN201110315458A CN102355422B CN 102355422 B CN102355422 B CN 102355422B CN 201110315458 A CN201110315458 A CN 201110315458A CN 102355422 B CN102355422 B CN 102355422B
- Authority
- CN
- China
- Prior art keywords
- packet
- dynamic channel
- passage
- shared
- dynamic
- 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.)
- Active
Links
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a multicore, parallel and lock-free quality of service (QOS) flow control method. The method comprises the following steps of: S1, receiving a data packet; S2, identifying the data packet; and S3, processing the data packet by a dispatcher of a QOS flow control system. Compared with the prior art, the method has the advantage that: the flexible, efficient and reliable bandwidth control capacity based on division of a business application priority and a multistage and multilevel bandwidth channel can be realized.
Description
Technical field
The present invention relates to the QOS flow control methods of a kind of multi-core parallel concurrent without lock, relate in particular to a kind of QOS flow control new method and system that realizes efficient each CPU parallel computation operation, nothing lock on multi-CPU system platform.
Background technology
Along with going deep into of informationization technology and developing rapidly of the Internet, All Around The World promptly combines together, and computer network has become national economic base and lifeblood.The network of oneself is being set up and developed in numerous enterprise, tissue and government departments all, and it is upper to be connected to Internet, fully to share, to utilize information and the resource of network.Be accompanied by the extensive use of the Internet, networking has become weighs a unit information, modern important symbol.In recent years, networking development was advanced by leaps and bounds, and constituent parts all drops into substantial contribution and built up in-house network and the Internet.Owing to lacking effective network traffic analysis and management control device, the development of network is faced with problems and challenge, has caused thus a series of safety, efficiency and legal issue.Unpredictable, performance bottleneck burst can affect the operational paradigm of crucial application.At present in most tissues, number of users is huge, network application environment is complicated, various P2P and Online Video application have taken a large amount of network bandwidths (as FLV, BitTorrent, Kazza, Emule etc.), the thing followed is, also day by day serious by the application performance decline problem of the congested initiation of network link, greatly affected and organized carrying out and the service quality of user's proper network application of regular traffic, network bandwidth resources is pernicious, and to take phenomenon serious.In addition, the operation of network key business and application can not get ensureing.So-called crucial application refers to the disparate networks application closely bound up with user's production and operation, information security, key user, comprises ERP, database, middleware, ecommerce, video conference etc.
In most tissues, all lack effective network management strategy at present, the application such as application (VOIP, video conference) to core business application (as ERP, database, OA, mail etc.), delay sensitive, instant messaging, P2P, Online Video, online game are made no exception, the use of network is not to arrange according to the priority of service application and significance level, finally causes network bandwidth resources to be preempted more severe such as online game, P2P, the immoderate use of Online Video application milli.
Because the use of the softwares such as P2P, online game, Online Video has taken the bandwidth resources of network at least 50%, and still having no to accelerate temperately engulfing valuable bandwidth resources, not only reduce the productivity ratio of enterprise, bring network congestion etc. to have a strong impact on the thorny problem that network uses, by having a strong impact on normal operation and the use of crucial application in tissue, finally can cause direct economic loss.
In China Internet develop rapidly, P2P and Online Video application are also popularized and are come rapidly, have also caused many troubles simultaneously.According to authoritative department statistics, current P2P and Online Video flow have accounted for approximately 70% of whole internet traffic, and with annual 350% speed increment.In network, P2P and Online Video traffic consumes the huge network bandwidth, improved network operation cost, network infrastructure be can't bear the heavy load.
Exploding of P2P and Online Video flow, can effectively control by the fluidic device being deployed on network.But therefore the network experience of normal network users but can not get ensureing, causes user to experience sharply and declines.
Summary of the invention
In order to address the above problem, the invention provides the QOS flow control methods of a kind of multi-core parallel concurrent without lock.
Correspondingly, a kind of multi-core parallel concurrent of the present invention comprises the following steps without the QOS flow control methods of lock:
Multi-core parallel concurrent, without a method for the QOS flow control of lock, comprises the following steps:
S1, reception packet;
S2, identify described packet;
S3, by packet described in the scheduler handle of QoS flow control system.
As a further improvement on the present invention, described S2 step comprises: by DPI & DFI, described packet is identified, and stamp corresponding VC and its from user's ID label.
As a further improvement on the present invention, described scheduler handle comprises queuing process and go out group process, and wherein, described queuing process is by join the team buffer memory or abandon of the packet of receiving; Described go out group process be responsible for finding out the packet that can send of buffer memory in system and give Rmon & Rmon2 resume module, until the packet of all buffer memorys is all sent or be can not find the packet that meets speed limit condition.
As a further improvement on the present invention, described queuing process specifically comprises the following steps:
S10, according to the User ID identifying on described packet and VC ID, to the dynamic channel of finding under corresponding VC passage under respective user;
If S20 can not find, create this passage, and check whether this VC passage exists Shared VC;
If S30 exists, confirm whether father's dynamic channel of sub-dynamic channel corresponding under Shared VC exists;
If S40 crosses and do not exist, need to create corresponding father's dynamic channel under Shared VC;
If S50 exists, increase the reference count of his father's dynamic channel; Confirm simultaneously and create each father's dynamic channel of correspondence under Pipe and Shared Pipe, thereby form a passage passage speed limit chain for each user, corresponding with the static speed limit chain of the VC/Shared VC/Pipe/Shared Pipe existing, collaborative work, and the pointer that the passage of Static and dynamic is pointed to parent channel by subchannel is linked to be a unidirectional chained list from bottom to top;
S60, sub-dynamic channel from the VC of the bottom start, and check step by step its token;
S70, negative if, carry out the filling of token according to single speed/2-rate 3-color token bucket algorithm, is still assignment if fill the token value of only having, and this packet of receiving done to discard processing, and queuing process finishes;
If the token inspection of the dynamic channel on all dynamic speed limit chain of S80 is all passed through, described packet is put into buffer queue and queues up, wait to be sent, and according to the length of this packet, deduct step by step the corresponding token value of each dynamic channel, queuing process finishes.
As a further improvement on the present invention, described in, going out group process comprises:
S100, check that the quantity of the packet of current whole QoS flow control system buffer memory is 1?
S101 is if enter fast path: from the queuing process of QoS scheduler, under VC passage, go out group packet;
If S102 can successfully go out packet of team, group process that goes out of QoS scheduler finishes at this point;
If S103 cannot successfully go out packet of team, enter slow-path,
Wherein, in described slow-path, what adopt is the mechanism that top-down poll is searched, the order that goes out team's traversal is from the most top Shared Pipe passage, travel through successively each subordinate subchannel, until all passages that exist in system are all traveled through one time or can successfully go out team's to packet below certain VC, finish this time to go out group process.
With respect to prior art, advantage of the present invention is: division, multistage multi-level bandwidth channel that can be based on service application priority, realize flexibly, efficient, bandwidth control ability reliably.
[brief description of the drawings]
Fig. 1 is the flow chart of an embodiment of the present invention multi-core parallel concurrent without the QOS flow control methods of lock;
Fig. 2 is the flow chart of an embodiment of the present invention scheduler queuing process;
Fig. 3 is the flow chart that the scheduler of an embodiment of the present invention goes out group process.
[embodiment]
In order to make the object, technical solutions and advantages of the present invention clearer, describe the present invention below in conjunction with the drawings and specific embodiments.
In an embodiment of the present invention, multi-core parallel concurrent is without the QOS flow control system of lock, the deep packet inspection technical that is core by DPI and DFI are the deep stream behavior detection technique of core, can be accurate to detection and the control of the packet to each session, simultaneously in conjunction with PFQ technology, the double-colored token bucket algorithm of three speed, self-defined Pipe passage is provided, self-defined VC(Virtual Channel) passage, guarantee bandwidth arranges, maximum bandwidth limit, bandwidth is leased, the control of Per-IP bandwidth, the control of Per-User bandwidth, share bandwidth control, weight arranges and applies a series of optimizing application and the Bandwidth Management control functions such as priority, support Inbound and two-way Bandwidth Management and the control of Outbound.
Wherein, according to self size of outlet bandwidth and the service condition of flow, user can the virtual bandwidth channel of self-defining, as user can be based on own service application situation, define different Pipe and VC bandwidth channel, and the agreement of different priorities or service application are joined in different VC passages, reach multi-level bandwidth channel management; By based on application priority management function priority, multichannel, thereby in the environment that ensures to use in the various bandwidth of tissue, can realize effective utilization and the control of bandwidth.
Article one, virtual link can have independently VC configuration.VC is closely related with the application type of network packet, and one or several similar network application types can be divided into a VC, and VC is the node type of the bottom in system.
Pipe is the set that is defined in some VC on VC, is the concept of introducing in order to embody three-dimensional flow control, wishes to meet user the demand that certain several type of service is controlled and retrained simultaneously.A VC can only belong to a definite Pipe.
Shared VC be defined under Pipe, the set of some VC on VC.Be the flow control in order to embody intersection and the concept introduced, wish to carry out between certain several type of service the demand that the competition of bandwidth resources is used to meet user.A Shared VC also can only belong to a definite Pipe.
Shared Pipe is the set that is defined in the Pipe of the some on Pipe.It is the concept in order to embody higher level solid, cross-current control characteristic and introduce.Shared Pipe is the most top control node.
Wherein, Vc, Pipe, Shared Vc, Shared Pipe are referred to as " node ", and the nodal community of node comprises:
NONE: also referred to as Normal attribute, the bandwidth control that has the node of this attribute is overall, controls separately regardless of user or IP address; Total be exactly and the bandwidth of this Node configuration flow through this node all flows bandwidth, that is: all users under this node share this total bandwidth.
PER-HOST/PER-USER: the bandwidth control that has the node of this attribute is not overall, but point IP, point user, each IP or user are controlled separately.And this node need arrange the maximum bandwidth of node, and the bandwidth enjoyed of each IP of this node of flowing through or each user.If this node bandwidth maximum bandwidth is 100Mbps, PER-HOST is configured to 1Mbps, and so, under the prerequisite that is 100Mbps in this node total bandwidth, each main frame can have the bandwidth of 1Mbps.
Weight: only have can join below Shared type node just to have this attribute as the node type of child node.Therefore in 4 kinds of node types, only have two kinds of nodes of Pipe and Vc to have this attribute, and the node of only likeing father, like son is while being the node of Shared type, this attribute just works.Its essential implication is the bandwidth (ensureing bandwidth and maximum bandwidth) that represents that this node can be got on its father node.In the time that VC or Pipe join Shared VC or Shared Pipe, the bandwidth that VC and Pipe can get is decided by weight ratio.
Default channel: i.e. after system operation, in order to carry out the forwarding of packet, the passage automatically creating.Comprise Default Pipe and Default VC.Under each direction (In-bound/Out-bound) of every link, there is a Default Pipe, have a Default VC below each Default Pipe.In addition, below each Pipe, also there is a Default VC.
Default Pipe and Default VC and common Pipe/VC are similar, and the configuration parameters such as its bandwidth can be revised, but can not be deleted.
As shown in Figure 1, the mechanism operation that multi-core parallel concurrent in an embodiment of the present invention adopts bag to drive without the system of the QOS flow control of lock, QoS flow control system is as the important component part of whole system, its core is exactly to process accordingly flowing into each packet of native system, comprises the abandoning and selection and the transmission of buffer memory, bag of calculating, bag of identification, CAR (the Committed Access Rate) token of user and passage under packet.Wherein, this multi-core parallel concurrent comprises without the method for the QOS flow control of lock:
S1, reception packet;
S2, identify described packet; Preferably, each packet receiving from network interface card, identifies by DPI & DFI, stamp corresponding VC and its from user's ID label;
S3, by packet described in the scheduler handle of QoS flow control system.Preferably, QoS scheduler handle comprises queuing (Enqueue) process and goes out the process two large divisions of team (Dequeue), queuing process is responsible for join the team buffer memory or abandon of the packet of receiving, be responsible for finding out the packet that can send of buffer memory in system and give Rmon & Rmon2 module and go out group process, the statistics of being correlated with only has and is forwarded, as long as and can find qualified packet, to be sustained, until the packet of all buffer memorys is all sent or be can not find the packet that meets speed limit condition.
As shown in Figure 2, an embodiment of the present invention scheduler queuing process, carries out traffic management and control as example taking every user, comprises the following steps:
S10, according to the User ID identifying on this packet and VC ID, to the dynamic channel of finding under corresponding VC passage under respective user;
If S20 can not find, create this passage, and check whether this VC passage exists Shared VC;
If S30 exists, confirm whether father's dynamic channel of sub-dynamic channel corresponding under Shared VC exists;
If S40 crosses and do not exist, need to create corresponding father's dynamic channel under Shared VC;
If S50 exists, increase the reference count of his father's dynamic channel; In like manner, confirm and create each father's dynamic channel of correspondence under Pipe and Shared Pipe, thereby form a passage passage speed limit chain for each user, corresponding with the static speed limit chain of the VC/Shared VC/Pipe/Shared Pipe existing, (static speed limit chain does overall Control for Speed Limitation for all users in collaborative work; Dynamically speed limit chain does specific Control for Speed Limitation for each user), and the pointer that the passage of these Static and dynamics points to parent channel by subchannel is linked to be a unidirectional chained list from bottom to top;
S60, sub-dynamic channel from the VC of the bottom start, and check step by step its token;
S70, negative if, first carry out the filling of token according to single speed/2-rate 3-color token bucket algorithm, if filling the token value of only having is still assignment, current this user's of explanation flow has exceeded its speed limit, therefore, this packet of receiving is done to discard processing, and queuing process finishes at this point.
If the token inspection of the dynamic channel on all dynamic speed limit chain of S80 is all passed through, the flow that current this user is described does not exceed its speed limit, therefore, this packet of receiving can be accepted the buffer queue of the system of putting into and queue up, etc. to be sent, and, according to the length of this packet, deduct step by step the corresponding token value of each dynamic channel, queuing process finishes.
As shown in Figure 3, an embodiment of the present invention scheduler goes out group process, comprises the following steps:
S100, check that the quantity of the packet of current whole QoS flow control system buffer memory is 1?
S101 is if so just enter fast path: directly from the queuing process of QoS scheduler before, just entered to attempt team below that VC passage of bag and go out a bag;
If S102 can successfully go out team and go out a packet, group process that goes out of QoS scheduler finishes at this point;
If S103 cannot successfully go out team and go out a packet, enter slow-path.
In the slow-path that goes out team, what adopt is the mechanism that top-down poll is searched, the order that goes out team's traversal is from the most top Shared Pipe passage, travel through successively each subordinate subchannel, until all passages that exist in system are all traveled through one time or can successfully go out team's to packet below certain VC, finish this time to go out group process.
Find a passage at every turn, need to judge that its token is for just? if not for just, fill its token according to single speed/2-rate 3-color token bucket algorithm, if token is still negative after filling, show that current this passage has exceeded speed limit, abandon this passage, find next passage at the same level, if its token is eligible, further check whether its DRR (Deficit round robin) is that just (DRR is the mechanism that send opportunity is introduced that obtains in order to ensure that all passages can both be fair: any one passage, the byte summation of its packet that can send is continuously conditional, reach this limits value, system will automatically switch to the passage of another one peer, in anti-locking system, occur that certain passage is going out bag and other the passage unjust phenomenon in " hunger " state always)? if not for just, fill its DRR value, and be switched to next passage at the same level, continue to find qualified passage.
In the time finding VC passage, attempt to go out packet of team from this VC, if can successfully go out team's to packet, finish this time to go out group process.
If traveled through all passages in system, still can not successfully go out team and go out a packet, finish this time to go out group process.
In sum, with respect to prior art, advantage of the present invention is: division, multistage multi-level bandwidth channel that can be based on service application priority, realize flexibly, efficient, bandwidth control ability reliably.
Wherein, bandwidth control ability is comprised:
Utilize the bandwidth grading function of system, for certain application-specific or some key customer in tissue specify minimum reserved bandwidth, to ensure that user can obtain same Bandwidth Management service in different time sections, different network environment for use and network uses impression;
For application-specific or some client specify maximum utilized bandwidth, thus the valuable bandwidth resources of the restriction non-critical application immoderate consumption of milli, thus ensure the service quality of crucial application;
Based on the control that ensures bandwidth and maximum bandwidth, in conjunction with the control of application priority, high priority agreement can be leased the bandwidth of free time or low priority protocol channel, thereby ensures that bandwidth obtains using reasonably, efficiently and managing;
Can control by the dynamic subchannel parameter of certain traffic channel is set the each IP address/user's who enters this passage guarantee bandwidth and maximum bandwidth;
Webmaster personnel can be according to demand, can be by several Pipe(or VC) put into a sharing channel, and different weights is set, to reach diversified flow control object.
Can realize the nested function of multilevel bandwidth by the relevant subordinate relation of Pipe, VC and the setting of Share Rate Pool passage and weight.
For convenience of description, while describing above device, being divided into various unit with function describes respectively.Certainly, in the time implementing the application, the function of each unit can be realized in same or multiple software and/or hardware.
As seen through the above description of the embodiments, those skilled in the art can be well understood to the mode that the application can add essential general hardware platform by software and realizes.Based on such understanding, the part that the application's technical scheme contributes to prior art in essence in other words can embody with the form of software product, this computer software product can be stored in storage medium, as ROM/RAM, magnetic disc, CD etc., comprise that some instructions (can be personal computers in order to make a computer equipment, server, or the network equipment etc.) carry out the method described in some part of each execution mode of the application or execution mode.
Device embodiments described above is only schematic, the wherein said unit as separating component explanation can or can not be also physically to separate, the parts that show as unit can be or can not be also physical locations, can be positioned at a place, or also can be distributed in multiple network element.Can select according to the actual needs some or all of unit wherein to realize the object of present embodiment scheme.Those of ordinary skill in the art, in the situation that not paying creative work, are appreciated that and implement.
The application can be used in numerous general or special purpose computing system environment or configuration.For example: personal computer, server computer, handheld device or portable set, laptop device, multicomputer system, system based on microprocessor, set top box, programmable consumer-elcetronics devices, network PC, minicom, mainframe computer, the distributed computing environment (DCE) that comprises above any system or equipment etc.
The application can describe in the general context of computer executable instructions, for example program unit.Usually, program unit comprises and carries out particular task or realize routine, program, object, assembly, data structure of particular abstract data type etc.Also can in distributed computing environment (DCE), put into practice the application, in these distributed computing environment (DCE), be executed the task by the teleprocessing equipment being connected by communication network.In distributed computing environment (DCE), program unit can be arranged in the local and remote computer-readable storage medium including memory device.
Be to be understood that, although this specification is described according to execution mode, but be not that each execution mode only comprises an independently technical scheme, this narrating mode of specification is only for clarity sake, those skilled in the art should make specification as a whole, technical scheme in each execution mode also can, through appropriately combined, form other execution modes that it will be appreciated by those skilled in the art that.
Listed a series of detailed description is above only illustrating for feasibility execution mode of the present invention; they are not in order to limit the scope of the invention, all do not depart from the equivalent execution mode that skill spirit of the present invention does or change and all should be included in protection scope of the present invention within.
Claims (1)
1. multi-core parallel concurrent, without a method for the QOS flow control of lock, is characterized in that, said method comprising the steps of:
S1, reception packet;
S2, identify described packet;
S3, by packet described in the scheduler handle of QoS flow control system; Described S2 step comprises: by DPI (Deep Packet Inspection) & DFI (Dynamic Flow Inspection), described packet is identified, and stamp corresponding VC (Virtual Channel) and its from user's ID label;
Described scheduler handle comprises queuing process and go out group process, and wherein, described queuing process is by join the team buffer memory or abandon of the packet of receiving; Described go out group process be responsible for finding out the packet that can send of buffer memory in system and give Rmon & Rmon2 (Remote Monitoring MIBs) resume module, until the packet of all buffer memorys is all sent or be can not find the packet that meets speed limit condition;
Described queuing process specifically comprises the following steps:
S10, according to the User ID identifying on described packet and VC ID, to the dynamic channel of finding under corresponding VC passage under respective user;
If S20 can not find the dynamic channel under respective user, create this passage, and check whether this VC passage exists Shared VC;
If S30 exists the dynamic channel under respective user, confirm whether father's dynamic channel of sub-dynamic channel corresponding under Shared VC exists;
S40, if there is no father's dynamic channel of corresponding sub-dynamic channel under Shared VC, need to create corresponding father's dynamic channel under Shared VC;
If S50 exists father's dynamic channel of sub-dynamic channel corresponding under Shared VC, increase the reference count of his father's dynamic channel; Confirm simultaneously and create each father's dynamic channel of correspondence under Pipe and Shared Pipe, thereby form a passage passage speed limit chain for each user, corresponding with the static speed limit chain of the VC/Shared VC/Pipe/Shared Pipe existing, collaborative work, and the pointer that the passage of Static and dynamic is pointed to parent channel by subchannel is linked to be a unidirectional chained list from bottom to top;
S60, sub-dynamic channel from the VC of the bottom start, and check step by step its token;
S70, negative if, carry out the filling of token according to single speed/2-rate 3-color token bucket algorithm, if the token value after filling is still negative, this packet of receiving done to discard processing, and queuing process finishes;
If the token inspection of the dynamic channel on all dynamic speed limit chain of S80 is all passed through, described packet is put into buffer queue and queues up, wait to be sent, and according to the length of this packet, deduct step by step the corresponding token value of each dynamic channel, queuing process finishes;
Described go out group process comprise:
S100, check whether the quantity of the packet of current whole QoS flow control system buffer memory is 1;
S101 is if enter fast path: from the queuing process of QoS scheduler, under VC passage, go out group packet;
If S102 can successfully go out packet of team, group process that goes out of QoS scheduler finishes at this point;
If S103 cannot successfully go out packet of team, enter slow-path,
Wherein, in described slow-path, what adopt is the mechanism that top-down poll is searched, the order that goes out team's traversal is from the most top Shared Pipe passage, travel through successively each subordinate subchannel, until all passages that exist in system are all traveled through one time or can successfully go out packet of team below certain VC, finish this time to go out group process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110315458.2A CN102355422B (en) | 2011-10-17 | 2011-10-17 | Multicore, parallel and lock-free quality of service (QOS) flow control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110315458.2A CN102355422B (en) | 2011-10-17 | 2011-10-17 | Multicore, parallel and lock-free quality of service (QOS) flow control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102355422A CN102355422A (en) | 2012-02-15 |
| CN102355422B true CN102355422B (en) | 2014-07-23 |
Family
ID=45578919
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110315458.2A Active CN102355422B (en) | 2011-10-17 | 2011-10-17 | Multicore, parallel and lock-free quality of service (QOS) flow control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102355422B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103634227B (en) * | 2012-08-20 | 2018-05-01 | 百度在线网络技术（北京）有限公司 | A kind of accurate dispatching method of service traffics based on number of users and device |
| CN104363115B (en) * | 2014-10-30 | 2017-11-03 | 任子行网络技术股份有限公司 | A kind of DPI is without lock stream mode management method and system |
| CN106230612B (en) * | 2016-07-12 | 2019-09-06 | 杭州迪普科技股份有限公司 | Handle the method and device of message |
| US10659372B2 (en) | 2017-01-25 | 2020-05-19 | Futurewei Technologies, Inc. | Multi-core lock-free rate limiting apparatus and method |
| CN109840145B (en) * | 2019-01-08 | 2021-09-07 | 苏州盛科通信股份有限公司 | Multistage scheduling method, device, network chip and storage medium |
| CN112817530B (en) * | 2021-01-22 | 2024-06-07 | 万得信息技术股份有限公司 | Method for reading and writing ordered data in full high efficiency through multiple lines Cheng An |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101060489A (en) * | 2007-05-16 | 2007-10-24 | 杭州华三通信技术有限公司 | Message forwarding method and device |
| CN101083563A (en) * | 2007-07-20 | 2007-12-05 | 杭州华三通信技术有限公司 | Method and apparatus for preventing distributed refuse service attack |
| CN101964754A (en) * | 2010-11-01 | 2011-02-02 | 南京邮电大学 | Point to point (P2P) service identification-based quality of service (QoS) routing method |
| CN102025623A (en) * | 2010-12-07 | 2011-04-20 | 苏州迈科网络安全技术股份有限公司 | Intelligent network flow control method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2371174A (en) * | 2001-01-11 | 2002-07-17 | Ericsson Telefon Ab L M | Controlling packet data flows in a telecommunications network |
-
2011
- 2011-10-17 CN CN201110315458.2A patent/CN102355422B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101060489A (en) * | 2007-05-16 | 2007-10-24 | 杭州华三通信技术有限公司 | Message forwarding method and device |
| CN101083563A (en) * | 2007-07-20 | 2007-12-05 | 杭州华三通信技术有限公司 | Method and apparatus for preventing distributed refuse service attack |
| CN101964754A (en) * | 2010-11-01 | 2011-02-02 | 南京邮电大学 | Point to point (P2P) service identification-based quality of service (QoS) routing method |
| CN102025623A (en) * | 2010-12-07 | 2011-04-20 | 苏州迈科网络安全技术股份有限公司 | Intelligent network flow control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102355422A (en) | 2012-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Noormohammadpour et al. | Datacenter traffic control: Understanding techniques and tradeoffs | |
| EP2865147B1 (en) | Guarantee of predictable and quantifiable network performance | |
| CN102355422B (en) | Multicore, parallel and lock-free quality of service (QOS) flow control method | |
| US9092269B2 (en) | Offloading virtual machine flows to physical queues | |
| CN102368741A (en) | Method supporting hierarchical queue scheduling and flow shaping and apparatus thereof | |
| CN1989738A (en) | Propagation of minimum guaranteed scheduling rates | |
| RU2643666C2 (en) | Method and device to control virtual output queue authorization and also computer storage media | |
| Li et al. | Finedge: A dynamic cost-efficient edge resource management platform for NFV network | |
| Kim et al. | An energy-aware service function chaining and reconfiguration algorithm in NFV | |
| Achar | Cloud-based System Design | |
| Liang et al. | Low-latency service function chain migration in edge-core networks based on open Jackson networks | |
| CN102947798B (en) | Computer system and control method thereof | |
| CN113938930B (en) | Construction method of virtual network function forwarding graph adapting to 5G network multi-service scene | |
| CN101557346B (en) | Polling-based packet queue output scheduling method and packet switching node | |
| CN102611924A (en) | Flow control method and system of video cloud platform | |
| CN111970149A (en) | Shared bandwidth realizing method based on hardware firewall QOS | |
| Sedaghat et al. | R2T-DSDN: reliable real-time distributed controller-based SDN | |
| Wang et al. | Service routing in multi-tier edge computing: A matching game approach | |
| Szymanski | Low latency energy efficient communications in global-scale cloud computing systems | |
| CN107332786B (en) | A kind of dispatching method ensureing data flow deadline under service chaining environment | |
| Moallemi et al. | On the flow-level dynamics of a packet-switched network | |
| Han et al. | A dynamic message-aware communication scheduler for Ceph storage system | |
| Kumari et al. | Study of traffic based load balancing algorithm in the SDN | |
| Zhao et al. | Joint Request Updating and Elastic Resource Provisioning With QoS Guarantee in Clouds | |
| Ma et al. | A link load balancing algorithm based on ant colony optimization in data center network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder | ||
| CP02 | Change in the address of a patent holder |
Address after: Room 301-302, 3rd Floor, Tiancheng Information Building, No. 88 South Tiancheng Road, High Speed Rail New City, Xiangcheng District, Suzhou City, Jiangsu Province, 215133 Patentee after: SUZHOU MAXNET NETWORK SAFETY TECHNOLOGY Co.,Ltd. Address before: 215021 International Science and Technology Park Phase III 8B, No. 1355 Jinjihu Avenue, Industrial Park, Suzhou City, Jiangsu Province Patentee before: SUZHOU MAXNET NETWORK SAFETY TECHNOLOGY Co.,Ltd. |