CN100433711C - Message speed limit method - Google Patents
Message speed limit method Download PDFInfo
- Publication number
- CN100433711C CN100433711C CNB2005100751285A CN200510075128A CN100433711C CN 100433711 C CN100433711 C CN 100433711C CN B2005100751285 A CNB2005100751285 A CN B2005100751285A CN 200510075128 A CN200510075128 A CN 200510075128A CN 100433711 C CN100433711 C CN 100433711C
- Authority
- CN
- China
- Prior art keywords
- message
- time window
- current time
- time
- cpu
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention relates to a message speed limiting method which comprises the steps that A. the width of time windows and the number of messages allowed to pass are preset; B. the time of the current system is acquired, and the passed time of a current time window is calculated according to the stop time of a previous time window; C. the passed time of the current time window is judged whether smaller than the width of time windows, if not, messages in the current time window are allowed to pass, and the stop time of the current time window is updated and ended, or the step D is executed; D. the number of the passed messages is judged whether larger than the number of messages allowed to pass, if is, the number of messages exceeding the passed messages is calculated, and the messages are allowed to pass after treated by flow quantity delay, or the messages in the current time window are allowed to pass, and the stop time of the current time window is updated. The present invention has the advantages that the use of CPU resources is optimized by flow quantity speed limit or flow quantity shaping for over-passed messages, CPU is protected when the system is attacked by broadcast storms or hackers, etc.
Description
Technical field
The present invention relates to the processing technology field of message, particularly relate to a kind of message rate-limiting method.
Background technology
Along with the communication chip technology rapid development, comprise that at present the data communications equipment processing speed of router, Ethernet switch etc. is more and more faster, port density is also done bigger and bigger, and function, the protocol characteristic of support are also more and more abundanter.Need variety of protocol message that the CPU of equipment handles, interruption, message etc. also more and more.And these protocol massages are after being received by CPU, to resolve on the one hand, and the message processing queue that is put into corresponding software protocol processing unit then is medium pending; On the other hand, present CPU processing speed is also more and more faster, also occurs on the equipment that has because send the message speed of protocol stack too fast and cause the message processing queue of some protocol processing unit to be overflowed and problems such as EMS memory occupation is too many in the short time on the CPU.
In the said equipment, particularly low end router, Ethernet switch equipment etc. generally are single CPU frameworks, CPU is except handling above-mentioned task, also need to handle a lot of important work: such as input, the output of the capable serial ports of response command, carry out corresponding order, veneer operating system to equipment is carried out task scheduling and maintenance, state to hardware forwarding chip and ancillary equipment is safeguarded, various memory read/writes such as internal memory and FLASH, SRAM etc. also may carry out message by CPU and transmit concerning router.Big in network traffics like this, the protocol characteristic of operation CPU more for a long time just often is in high load capacity even full load condition, when receiving message more in this case, can occur that serial ports low-response, protocol massages are lost, dynamic routing is lost and dynamic address resolution agreement (ARP, Address Resolution Protocol) such as loses at problem.How under heavy traffic condition, especially when network exists broadcast storm or equipment to be attacked, effectively utilize cpu resource, the assurance system is stable to be the major issue that Ethernet switch equipment need solve.
From at present to the CPU working condition analysis of Ethernet switch equipment as can be known, it is to optimize one of main means that cpu resource utilizes that CPU is received that message limits.Because there are many messages all to require transmitted to CPU to handle, but wherein some message for example the ARP broadcasting packet can be repeated to deliver to CPU, and in fact to the foundation of the mapping relations of an IP address to MAC address, CPU only need receive that an ARP message gets final product.And often a lot of Ethernet switch equipment be because a variety of causes all will broadcast and the multicast message transmitted to CPU, and this has on network, and the CPU to equipment causes very big pressure in the broadcast storm.
At present, the control method for message mainly contains three kinds.
First kind is the method that CPU receives the message restriction, this method is to do some controls from the forwarding chip hardware designs, the IXE2412 chip of Intel Company for example, to IGMP (IGMP, InternetGroup Management Protocol), the ARP message all has independent queue management, guarantee that by register setting these messages can access priority treatment, general broadcasting, multicast message are then given lower priority, when CPU is busy with one's work, the low priority message is carried out random drop.All do not provide the controlled function of the CPU packet receiving classification on this hardware but problem is at present a lot of ASIC forwarding chips, the controlled function that perhaps provides is limited, can't satisfy the needs that protocol characteristic is enriched in the support of present Ethernet switch equipment.
Second kind is to adopt the method for traffic classification to control the flow of message.Described method limits various messages transmitted to CPU and the bandwidth assurance by access control lists ACL is set.But based on the reason of above-mentioned first method, support more protocol characteristic need add tens even more ACL list item, this takies too big to the hardware flow classified resource; And be subjected to the traffic classification limited processing ability (quantity, the field location of the traffic classification keyword that extracts that can support are limited) of some forwarding chips from message, also some protocol massages can't be discerned and control by traffic classification.
The third is to adopt software approach that the message of transmitted to CPU is limited with bandwidth to guarantee.Described employing software approach mainly is to adopt timer to reporting the CPU message to limit to be, and the operation that comes supervision timer with special unit.Described method comprises: at first, write the call back function of timer according to user's needs; Secondly, to system's application hardware timer resource, if the application failure then finishes; Once more, the overtime call back function of the preset timer time interval and timer starts timer; At last, timer moves automatically, and each call back function that just calls setting when overtime is automatically done corresponding message rate-limiting or traffic shaping operation.
As the above analysis, the shortcoming of prior art is: 1) can not effectively utilize cpu resource; 2) need take the timer resource of hardware, and the overtime call back function of preset timer need consume a large amount of software resources, and increase the complexity of operating system, increase the workload and the risk of operating system task scheduling.Can not effectively protect when 3) suffering broadcast storm or assault CPU in system.
Summary of the invention
The technical problem that the present invention solves provides a kind of message rate-limiting method; when this method surpasses predetermined message amount to solve the message that receives at CPU; by the message that exceeds standard is carried out Limit Rate or traffic shaping, thus reach optimize cpu resource use, when suffering broadcast storm or assault in system to problems such as CPU protect.
For addressing the above problem, the invention provides a kind of message rate-limiting method, described method comprises step:
Allow the message amount passed through in A1, Preset Time window width and this time window;
B1, obtain current system time, and spend the time according to what previous time window calculated the current time window deadline;
C1, judge the current time window the time that crosses whether less than the width of time window, if not, then allow the message in the current time window to pass through, and upgrade the cut-off time of current time window, finish; If, execution in step D1;
D1, whether judge message amount that the current time window passed through, if calculate to exceed and allow the message amount passed through, and it is carried out allowing to pass through after traffic delay is handled greater than the message amount that allows to pass through; Otherwise the message that allows to send in the normal reception of current time window is until reaching the default quantity of passing through.
This method is embedded in the transmitting-receiving message task.
Before steps A 1, described processor carries out filtration treatment to the message that receives.
Calculate described in the step D1 to exceed and allow the message amount passed through, and it is carried out implementation procedure that traffic delay handles comprise:
41) allow the difference of the message amount passed through to calculate to exceed to allow the message amount passed through in the message amount that receives by the current time window and the default unit interval window;
42) calculate according to the message rate that allows in the unit interval window to pass through and exceed the time that the message needs that allow to pass through postpone;
43) exceed the time that message that permission passes through postpones according to described needs and postpone to handle described;
44) when arriving time of delay, allow the described message that allows to pass through that exceeds to pass through, and upgrade the cut-off time of current time window.
Step 42) calculating the computing formula exceed the time that the message that allows to pass through need postpone described in is: needs postpone time=(surpass allow in the message amount-unit interval window that allows to pass through in the default unit interval window pass through message amount)/unit interval allows the quantity * time window width of reception message.
In addition, the present invention also provides a kind of message rate-limiting method, and accompanying method comprises step:
Allow the message amount passed through in A2, Preset Time window width and the time window;
B2, obtain current system time, and spend the time according to what previous time window calculated the current time window deadline;
C2, judge the current time window the time that crosses whether less than the time window width, if not, then allow the message in the current time window to pass through, and upgrade the cut-off time of current time window, finish; If, execution in step D2;
D2, judge the current time window message amount whether passed through of the quantity by message greater than default permission, if, directly abandon the message that exceeds standard, finish; Otherwise the message that allows to send in the normal reception of current time window is until reaching the default quantity of passing through.
This method is embedded in the transmitting-receiving message task.
Before steps A 2, described processor carries out filtration treatment to the message that receives.
Compared with prior art, the present invention has following beneficial effect: the present invention comes the transmitted to CPU message is carried out flow restriction or traffic shaping by the quantity of passing through in time window width and the unit interval window is set, quantity with the literary composition of delivering newspaper on guaranteeing in the unit interval window is basicly stable, thereby guarantee that effectively CPU reduces the utilization rate of CPU under the impact of big flow message, do not influence the normal operation of variety of protocol function simultaneously again.The present invention and prior art are compared by the software limitations mode of creating timer and transmitting-receiving message task, have realize easy, to take resource few, speed limit time window, maximum stream flow threshold value etc. is easy in actual motion advantages such as real-time adjustment.Particularly on switching equipment (such as Ethernet switch), used method of the present invention at certain; the civilian quantity of delivering newspaper on not only guaranteeing in the unit interval window is basicly stable; and can also realize protection preferably to CPU; reduce CPU usage significantly, do not influenced the normal operation of variety of protocol function simultaneously.In addition, when cpu load is heavier, preferably divide formation to operate, just each message is forwarded to the hardware queue of CPU, and in individual queue, set different threshold values, and send earlier to guarantee the higher important message of priority, be dropped to avoid important message.
Description of drawings
Fig. 1 is a kind of flow chart of message rate-limiting method of the present invention;
Fig. 2 is the another kind of flow chart of message rate-limiting method of the present invention;
Fig. 3 is the optimization flow chart of message rate-limiting method of the present invention.
Embodiment
Core of the present invention is that the flow by message comes Controlled CPU to receive the flow of message in the time window by being provided with; and take corresponding penalty mechanism to exceeding in the unit interval window message; promptly postpone to handle; basicly stable with the message amount of sending in guarantor unit's time window, thus the purpose of CPU being protected when reaching the use of optimizing cpu resource and suffering broadcast storm or assault in system.
The invention provides a kind of general easy message rate-limiting method, this method by software approach on the CPU message that send carry out Limit Rate or traffic shaping, be applicable to various types of CPU and operation system of software.Below with the technical scheme that example is further told about this invention that is embodied as on embedded OS Vxworks.
Please refer to Fig. 1, flow chart for message rate-limiting method of the present invention, this method is in control procedure, can move one during the Ethernet switch equipment operation and be responsible for the task of receiving message specially, elder generation of the present invention is embedded into corresponding code in this transmitting-receiving message task handling function, obtains handling in each message receiving course.Described method specifically comprises:
Step S10: allow quantity in the width of Preset Time window and the time window by message;
Step S11: obtain current system time, and calculate deadline according to previous time window the time excessively of current time window;
Step S12: whether the time that crosses of judging the current time window if not, then allows the message in the current time window to pass through, and upgrades the cut-off time of current time window less than the width of time window, finishes (step S13); If carry out (step S14);
Step S14: judge the current time window the quantity by message whether greater than the quantity that allows by message, if calculate to exceed and allow the message amount passed through, and described message is carried out punitive postpone to handle the back and allow by (step S15); Otherwise, allow the message in the current time window to pass through, and upgrade the cut-off time (step S13) of current time window.
Before step S10, the message that transmitting-receiving message task receives transmitted to CPU carries out filtration treatment to the message that reports CPU, and this is because some message may be that system is uninterested, perhaps some agreement, functional module require its message that abandons is then filtered out it in advance.For example, if system does not start OSPF (OSPF, Open Short Path First), described OSPF is a kind of network routing protocol.In this step, if receive the ospf protocol message, just can directly it be filtered out for the reception task, promptly directly abandon in this step, do not do next step processing; Can just limit the flow control of part of message like this at transmitting-receiving message task place.
In step S10, the processing function that provides by operating system obtains current system time, and it is kept at variable ulNowTicks.Such as, the VxWorks system can use the sysTimeBaseIn function.Embed the corresponding code of the present invention in the described processing function in advance, so that being handled accordingly in each message receiving course.Time that crosses of current time window described in this step by current system time with on the difference of cut-off time time calculate.Its computing formula is: the current time window has been crossed time window cut-off time ulOldTicks on the current system time ulNowTicks one of time ulTicksGo=, and wherein, described time window cut-off time is meant the moment of being write down during by time window when message.When initial start, a last time window is by constantly being initialized as 0.The described moment all has Tick to count.When the current time window receives message, and its message amount PacketsReceived added one, wherein, be preset as 0 during described PacketsReceived initialization.
In step S12, judge whether the current time window is overtime, whether the time ulTicksGo excessively that is the current time window is less than preset time window width ulTimeWindow, wherein, it is default when described time window width is initialization, be provided with according to systematic function and demand, and be kept in the global variable, being in operation can be according to the actual conditions adjustment.
If the current time window is overtime, the time that crosses that is to say the current time window is during less than the preset time window width, then the current time window does not finish, and judges whether the current time window has passed through the quantity PacketsPermit (step S14) of message greater than permission by the quantity PacketsReceived of message again; If; the quantity that the normal message up sending that receives in this time window then is described exceeds standard; need take the flow control measure; that is to say; when having surpassed initialization, the message amount that receives in the current time window allows the message amount PacketsPermit that passes through in the default unit interval window; in order to optimize the utilization of resources of CPU; the present invention to the message that exceeds standard by penalty mechanism to control its flow, the message amount of sending on guaranteeing in average unit interval window basicly stable (step S15).The message amount that allows in the wherein said unit interval window to pass through is to be provided with according to systematic function and demand, and be kept in the global variable, being in operation can be according to the actual conditions adjustment, calculating exceeds the message number of permission, and its computing formula is: the message amount that receives in the current unit interval window is the message amount that exceeds standard in the unit interval window with the difference of the quantity that the interior permission of default unit interval window is passed through.
Wherein, described penalty mechanism needs that just the message that exceeds standard that is calculated is carried out certain flow and postpones to handle, i.e. " punitive " time-delay is to reach the effect of the reception shaping processing on the time shaft.The implementation procedure that described punitive postpones comprises:
A. allow the difference of the message amount passed through to calculate to exceed to allow the message amount passed through in the message amount that receives by the current time window and the default unit interval window;
B. calculate according to the message rate that allows in the unit interval window to pass through and exceed the time that the message needs that allow to pass through postpone;
C. exceed the time that message that permission passes through postpones according to described needs and postpone to handle described;
D. when arriving time of delay, allow the described message that allows to pass through that exceeds to pass through, and upgrade the cut-off time of current time window.
That is to say that it is to calculate the time that needs are delayed time according to allowing by the message Mean Speed that described punishment postpones.And that described penalty mechanism can have is multiple; the processing that the present invention realizes at present is: if message up sending speed is higher; just surpassed in unit interval and to have allowed the message amount PacketsPermit that passes through in the default unit interval window less than time window; that is just according to the message amount PacketsPermit that allows in the message amount PacketsReceived-unit interval window that surpasses to pass through; and then divided by allowing to receive message amount PacketsPermit in the unit interval window; multiply by preset time window width ulTimeWindow at last again; needing just to obtain the time of time-delay, this time also is to count with Tick.But the present invention is not limited to described penalty mechanism, can also realize by other mode.In addition, current message is carried out buffer memory, and, current message is made delay process according to the delay value that is calculated; Allow current message to pass through during the moment when arriving delay, be about to this message up sending and handle to protocol stack.
When the time that the current time window has been crossed has been not less than time width, then finish the current time window, allow current message to pass through, and the message amount zero clearing (step S13) that the current time window is received.As window cut-off time new current time, promptly ulOldTicks equals ulNowTicks with current time ulNowTicks.Begin next cycle of treatment then.
The present invention guarantees to allow in the unit interval window message rate of transmitted to CPU to be limited in the regular hour scope by said process, if the transmitted to CPU message rate exceeds standard in short-term, then take corresponding " punishment " to postpone mechanism, basicly stable with the message amount that guarantees average unit interval window transmitted to CPU.In transmitted to CPU message Mean Speed not quite but when big peak rate was arranged once in a while, method of the present invention can realize the protection of CPU effectively and can guarantee not lose important message.Operating system can be dispatched CPU and be handled other task in the time-delay process.
In addition, the present invention also provides another kind of message rate-limiting method, and its flow chart sees Fig. 2 for details, and described method comprises step:
Step H10: allow quantity in the width of Preset Time window and the time window by message;
Step H11: obtain current system time, and calculate deadline according to previous time window the time excessively of current time window;
Step H12: whether the time that crosses of judging the current time window if not, then allows the message in the current time window to pass through, and upgrades the cut-off time of current time window less than the width of time window, finishes (step H13); If carry out (step H14);
Step H14: judge the current time window the quantity by message whether greater than the quantity of default permission by message, if, directly abandon the message that exceeds standard, finish (step H15); Otherwise the message that allows to send in the normal reception of current time window is until reaching the default quantity of passing through (step H16).
This method of the present invention also can guarantee and will allow the message rate of transmitted to CPU to be limited in certain scope in the unit interval window.Described this method mainly is when detecting the transmitted to CPU message and exceeded standard, and directly abandons this message.Though may packet loss in implementation procedure,, this method is various in the equipment operation agreement, can effectively realize the protection to CPU when cpu load is heavier.
This message rate-limiting method of the present invention is the simplification to first kind of message rate-limiting method, and the performing step part of implementation step and first method is similar, is specially:
Before step H10, the message that transmitting-receiving message task receives transmitted to CPU carries out filtration treatment to the message that reports CPU, and this is because some message may be that system is uninterested, perhaps some agreement, functional module require its message that abandons is then filtered out it in advance.For example, if system does not start OSPF (OSPF, Open Short Path First), described OSPF is a kind of network routing protocol.In this step, if receive the ospf protocol message, just can directly it be filtered out for the reception task, promptly directly abandon in this step, do not do next step processing; Can just limit the flow control of part of message like this at transmitting-receiving message task place.
In step H10, the processing function that provides by operating system obtains current system time, and it is kept at variable ulNowTicks.Such as, the VxWorks system can use the sysTimeBaseIn function.Embed the corresponding code of the present invention in the described processing function in advance, so that being handled accordingly in each message receiving course.Time that crosses of current time window described in this step by current system time with on the difference of cut-off time time calculate.Its computing formula is: it is that current system time ulNowTicks deducts a time window cut-off time ulOldTicks that the current time window has been crossed time ulTicksGo, and wherein, described time window cut-off time is meant the moment of being write down during by time window when message.When initial start, a last time window is by constantly being initialized as 0.The described moment all has Tick to count.When the current time window receives message, and its message amount PacketsReceived added one, wherein, be preset as 0 during described PacketsReceived initialization.
In step H12, judge whether the current time window is overtime, whether the time ulTicksGo excessively that is the current time window is less than preset time window width ulTimeWindow, wherein, it is default when described time window width is initialization, be provided with according to systematic function and demand, and be kept in the global variable, being in operation can be according to the actual conditions adjustment.If the time that crosses of current time window is less than the preset time window width, then in the current time window end is arranged, judge again the current time window the quantity by message whether greater than the quantity that allows to pass through, if do not surpass, then the normal quantity that receives message in this time window is sent on normal, send on just stopping until reaching predetermined limit, and finish this processing; If surpass, illustrate that then the normal quantity that receives message exceeds standard in this time window, directly abandon the described message that exceeds standard.
If the time that crosses of current time window is not less than the preset time window width, then allow the message in the current time window to pass through, and finish the current time window and the quantity zero clearing of current time window, and ulOldTicks deadline that upgrades the current time window.
The described process of this method also can effectively guarantee to allow the message rate of transmitted to CPU to be limited in certain scope in the unit interval window.Message is directly abandoned if the interior transmitted to CPU message of unit interval window exceeds standard then.This method may packet loss, but can also realize the protection to CPU well because various in the equipment operation agreement, support characteristic abundant, more suitable when cpu load is heavier.
In addition, the present invention provides a kind of message rate-limiting method again, this method for cpu load when heavier for optimizing message rate-limiting method, its flow chart sees Fig. 3 for details.This method adopts and divides formation to operate, and set different threshold values in individual queue when cpu load is heavier, send earlier to guarantee the higher important message of priority, is not dropped substantially to guarantee important message.Wherein, the described hardware queue that just each message is forwarded to CPU with the implementation of dividing formation; all adopt the method for message flow control as described in Figure 3; this method can be set bigger unit interval message to high-priority queue and pass through quantity; Low Priority Queuing is set less unit interval message pass through quantity; when realizing CPU protected, to the grading control and the service of different business.
The optimization flow process of message rate-limiting method as shown in Figure 3 comprises step:
Step M10: allow the message amount passed through in the time window width of default respectively each formation and this time window respectively;
Step M11: obtain current system time, and calculate deadline according to previous time window the time excessively of current time window;
Step M12: whether the time that crosses of judging the current time window respectively if not, then allows the message in the current time window to pass through, and upgrades the cut-off time of current time window less than the width of time window, finishes (step M13); If, execution in step M14;
Step M14:, whether the message amount of judging each formation that the current time window has passed through respectively allow the message amount passed through greater than each formation respectively, if, calculate respectively to exceed and allow the message amount passed through, and it is postponed to handle the back allow by (step M15); Otherwise, allow the message in each current time window to pass through, and upgrade the cut-off time (step M13) of current time window.
The something in common of described method and said method, here repeat no more, its difference is: earlier the message that receives is divided into each different formations according to the type or the priority of message, and then preset according to the difference of each queue priority and to allow the message amount passed through in the unit interval window, high-priority queue is set the message amount of passing through in the bigger unit interval window, Low Priority Queuing is set the message amount of passing through in the less unit interval window, thereby during the message amount that the message amount of each formation of having passed through when the current time window allows to pass through greater than each formation, calculate earlier each formation and exceed and allow the message amount passed through, and then this is carried out traffic delay and handle the back and allow it to pass through.Wherein comprise: allow the difference of the message amount passed through to calculate each formation in the message amount that receives by each formation current time window and the default unit interval window and exceed and allow the message amount passed through for the specific implementation process of the data message that exceeds standard by flow control (be traffic delay handle); Calculate each formation according to the message rate that allows in each formation unit interval window to pass through and exceed the time that the message needs that allow to pass through postpone, its computing formula of calculating the time of each queue delay is: need to postpone time=(surpass allow in the message amount-unit interval window that allows to pass through in the default unit interval window pass through message amount)/unit interval allows to receive the quantity * time window width of message; Exceed the time that message that permission passes through postpones according to described needs and postpone respectively to handle described; When arriving time of delay, allow described each formation to exceed the message that allows to pass through and pass through, and upgrade the cut-off time of each formation current time window.
The above only is a preferred implementation of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (3)
1, a kind of message rate-limiting method is characterized in that, comprises step:
Allow the message amount passed through in A2, Preset Time window width and the time window;
B2, obtain current system time, and spend the time according to what previous time window calculated the current time window deadline;
C2, judge the current time window the time that crosses whether less than the time window width, if not, then allow the message in the current time window to pass through, and upgrade the cut-off time of current time window, finish; If, execution in step D2;
D2, judge the current time window message amount whether passed through of the quantity by message greater than default permission, if, directly abandon the message that exceeds standard, finish; Otherwise the message that allows to send in the normal reception of current time window is until reaching the default quantity of passing through.
According to the described message rate-limiting method of claim 1, it is characterized in that 2, this method is embedded in the transmitting-receiving message task.
According to the described message rate-limiting method of claim 2, it is characterized in that 3, before steps A 2, described transmitting-receiving message task comprises that the message to receiving carries out filtration treatment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100751285A CN100433711C (en) | 2005-06-08 | 2005-06-08 | Message speed limit method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100751285A CN100433711C (en) | 2005-06-08 | 2005-06-08 | Message speed limit method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1725732A CN1725732A (en) | 2006-01-25 |
| CN100433711C true CN100433711C (en) | 2008-11-12 |
Family
ID=35924981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100751285A Expired - Fee Related CN100433711C (en) | 2005-06-08 | 2005-06-08 | Message speed limit method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100433711C (en) |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101001249A (en) * | 2006-12-31 | 2007-07-18 | 华为技术有限公司 | Method and device for preventing IGMP message attack |
| CN101034975B (en) * | 2007-04-05 | 2010-05-26 | 华为技术有限公司 | Method and device for preventing small packet attack |
| CN101083563B (en) * | 2007-07-20 | 2010-08-11 | 杭州华三通信技术有限公司 | Method and apparatus for preventing distributed refuse service attack |
| CN101299719B (en) * | 2008-06-04 | 2012-07-25 | 北京星网锐捷网络技术有限公司 | Detection processing method for data flow, central processing unit and switch |
| CN101924679A (en) * | 2009-06-12 | 2010-12-22 | 中兴通讯股份有限公司 | Message rate-limiting method of Ethernet port, device and data communication device |
| CN101951339B (en) * | 2010-08-24 | 2012-08-15 | 北京星网锐捷网络技术有限公司 | Message speed limitation method, device and network equipment |
| CN102098215B (en) * | 2010-12-17 | 2014-03-12 | 曙光信息产业股份有限公司 | Priority management method for multi-application packet reception |
| US8599759B2 (en) * | 2011-04-29 | 2013-12-03 | Cooper Technologies Company | Multi-path radio transmission input/output devices, network, systems and methods with on demand, prioritized routing protocol |
| CN102710502B (en) * | 2012-05-12 | 2015-01-07 | 杭州迪普科技有限公司 | Network speed-limiting method and device based on time wasting |
| CN102882809B (en) * | 2012-10-26 | 2015-07-15 | 杭州迪普科技有限公司 | Network speed-limiting method and device based on message buffering |
| CN102938722B (en) * | 2012-10-30 | 2017-11-10 | 中兴通讯股份有限公司 | The method and device that a kind of network equipment periodic protocol is given out a contract for a project |
| CN103117871B (en) * | 2012-12-31 | 2016-08-31 | 广东东研网络科技股份有限公司 | Message method for limiting for OLT network management channel (nmc) |
| CN104038494A (en) * | 2014-06-11 | 2014-09-10 | 普联技术有限公司 | Method for recording attack source and exchanger |
| US9924398B2 (en) * | 2015-10-13 | 2018-03-20 | Quanta Computer Inc. | Method for reducing load by filtering out broadcast messages |
| CN105897365B (en) * | 2016-03-23 | 2019-01-18 | 京信通信系统（中国）有限公司 | A kind of protecting against shock processing method and processing device of processor |
| CN106230741A (en) * | 2016-09-18 | 2016-12-14 | 杭州迪普科技有限公司 | A kind of method and apparatus that message is carried out speed limit |
| EP3720069B1 (en) * | 2017-12-31 | 2025-02-19 | Huawei Technologies Co., Ltd. | Packet sending method and device |
| CN110852903B (en) * | 2019-11-15 | 2023-01-24 | 积成电子股份有限公司 | Ladder type elastic network storm filtering method based on FPGA |
| CN111526060B (en) * | 2020-06-16 | 2023-02-28 | 网易（杭州）网络有限公司 | Method and system for processing service log |
| CN111740870B (en) * | 2020-08-27 | 2021-02-02 | 苏州浪潮智能科技有限公司 | Network dynamic management method, system and device |
| CN112468413B (en) * | 2020-11-26 | 2023-04-25 | 迈普通信技术股份有限公司 | Message processing method and device, electronic equipment and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1525717A (en) * | 2003-02-25 | 2004-09-01 | ������������ʽ���� | Traffic shaping method and traffic shaping device |
| US20040174816A1 (en) * | 2003-03-07 | 2004-09-09 | Cisco Technology, Inc. | System and method for communicating data in a network environment |
| US20040221032A1 (en) * | 2003-05-01 | 2004-11-04 | Cisco Technology, Inc. | Methods and devices for regulating traffic on a network |
| CN1567828A (en) * | 2003-06-18 | 2005-01-19 | 中兴通讯股份有限公司 | A method for performing speed limiting on data traffic by network processor |
| CN1571538A (en) * | 2004-05-10 | 2005-01-26 | 中兴通讯股份有限公司 | A method for performing system load control on mobile switch center |
| WO2005032068A1 (en) * | 2003-09-29 | 2005-04-07 | British Telecommunications Public Limited Company | Bandwidth allocation |
| CN1606293A (en) * | 2003-10-07 | 2005-04-13 | 阿尔卡特公司 | Line card port protection rate limiter circuitry |
| WO2005039124A1 (en) * | 2003-10-17 | 2005-04-28 | Tellabs Oy | Method and equipment for performing flow shaping that maintains service quality in packet-switched telecommunications |
-
2005
- 2005-06-08 CN CNB2005100751285A patent/CN100433711C/en not_active Expired - Fee Related
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1525717A (en) * | 2003-02-25 | 2004-09-01 | ������������ʽ���� | Traffic shaping method and traffic shaping device |
| US20040174816A1 (en) * | 2003-03-07 | 2004-09-09 | Cisco Technology, Inc. | System and method for communicating data in a network environment |
| US20040221032A1 (en) * | 2003-05-01 | 2004-11-04 | Cisco Technology, Inc. | Methods and devices for regulating traffic on a network |
| CN1567828A (en) * | 2003-06-18 | 2005-01-19 | 中兴通讯股份有限公司 | A method for performing speed limiting on data traffic by network processor |
| WO2005032068A1 (en) * | 2003-09-29 | 2005-04-07 | British Telecommunications Public Limited Company | Bandwidth allocation |
| CN1606293A (en) * | 2003-10-07 | 2005-04-13 | 阿尔卡特公司 | Line card port protection rate limiter circuitry |
| WO2005039124A1 (en) * | 2003-10-17 | 2005-04-28 | Tellabs Oy | Method and equipment for performing flow shaping that maintains service quality in packet-switched telecommunications |
| CN1571538A (en) * | 2004-05-10 | 2005-01-26 | 中兴通讯股份有限公司 | A method for performing system load control on mobile switch center |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1725732A (en) | 2006-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100433711C (en) | Message speed limit method | |
| US7616573B2 (en) | Fair WRED for TCP UDP traffic mix | |
| Bux et al. | Flow control in local-area networks of interconnected token rings | |
| US8437252B2 (en) | Intelligent congestion feedback apparatus and method | |
| US6308228B1 (en) | System and method of adaptive message pipelining | |
| US10728156B2 (en) | Scalable, low latency, deep buffered switch architecture | |
| Siemon | Queueing in the Linux network stack | |
| CN108243117B (en) | Flow monitoring method and device and electronic equipment | |
| Hu et al. | TLB: Traffic-aware load balancing with adaptive granularity in data center networks | |
| US7480238B2 (en) | Dynamic packet training | |
| CN101924679A (en) | Message rate-limiting method of Ethernet port, device and data communication device | |
| WO2012068804A1 (en) | Message processing method and device | |
| US20070041319A1 (en) | Data packet rate control | |
| CN102299850A (en) | Method and device for protecting central processing unit (CPU) | |
| US20140207948A1 (en) | Network interface on a chip with an adaptive system to trigger data forwarding | |
| CN116170377B (en) | Data processing method and related equipment | |
| CN110601996B (en) | Looped network anti-starvation flow control method adopting token bottom-preserving distributed greedy algorithm | |
| Barczyk et al. | High rate packets transmission on Ethernet LAN using commodity hardware | |
| Madden | Challenges using the Linux network stack for real-time communication | |
| Löser et al. | Real Time on Ethernet using off-the-shelf Hardware | |
| Luangsomboon et al. | Necessary and Sufficient Condition for Triggering ECN Before PFC in Shared Memory Switches | |
| Biswas et al. | A High Performance Packet Capturing Support for Alarm Management Systems. | |
| CN103095646B (en) | A kind of method of reseptance of network data | |
| Neufeld et al. | The LHCb eventbuilder: Design, implementation and operational experience | |
| CN100426789C (en) | High-speed message transmitting method with high reliability |
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 | ||
| CP03 | Change of name, title or address |
Address after: 310052 Binjiang District Changhe Road, Zhejiang, China, No. 466, No. Patentee after: NEW H3C TECHNOLOGIES Co.,Ltd. Address before: 310053 Hangzhou hi tech Industrial Development Zone, Zhejiang province science and Technology Industrial Park, No. 310 and No. six road, HUAWEI, Hangzhou production base Patentee before: HANGZHOU H3C TECHNOLOGIES Co.,Ltd. |
|
| CP03 | Change of name, title or address | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20081112 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |