CN101984594A - Switching network flow control method and device - Google Patents
Switching network flow control method and device Download PDFInfo
- Publication number
- CN101984594A CN101984594A CN2010105273825A CN201010527382A CN101984594A CN 101984594 A CN101984594 A CN 101984594A CN 2010105273825 A CN2010105273825 A CN 2010105273825A CN 201010527382 A CN201010527382 A CN 201010527382A CN 101984594 A CN101984594 A CN 101984594A
- Authority
- CN
- China
- Prior art keywords
- ply
- yarn drill
- data
- buffer memory
- descending
- 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.)
- Pending
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The embodiment of the invention discloses a switching network flow control method which comprises the following steps: receiving switching network data and judging the magnitude relations between the cache occupation of the received data and a lower threshold and between the cache occupation of the received data and an upper threshold by a downlink line card, wherein the upper threshold is greater than the lower threshold; when the cache occupation is greater than the lower threshold, sending a flow control signal to an uplink line card, wherein the flow control signal is used for enabling the uplink line card to reduce the data sending rate; and when the cache occupation is greater than the upper threshold, sending a back pressure signal to a superior exchange chip connected with the downlink line card to enable the superior exchange chip to stop sending data. The embodiment of the invention also provides a corresponding device which comprises line cards and a router. By adopting the technical scheme of the invention, when the cache occupation of the downlink line card exceeds the lower threshold, the data sending rate of the uplink line card is adjusted, and the cache occupation of the downlink line card can be controlled below a certain level as much as possible, thereby avoiding the problems of deadlock and vicious circle caused by that the downlink line card sends the back pressure signal to the superior chip.
Description
Technical field
The present invention relates to communication technical field, be specifically related to a kind of switching network flow control methods and device.
Background technology
The Internet is mainly by switching node with the corresponding communication circuit is interconnected forms, and wherein most important switching node is a router.Router mainly is made up of ply-yarn drill and switching fabric.According to the trend of data traffic, ply-yarn drill can be divided into up ply-yarn drill and descending ply-yarn drill.Switching fabric is the bridge that connects data-in port and output port, and it provides data channel for ply-yarn drill, makes packet can be forwarded to the purpose ply-yarn drill from the ply-yarn drill that enters.Switching fabric can be single-stage, back-to-back with multistage.
Data exchange ways comprises two kinds in the switching network: the port that the data traffic in the up ply-yarn drill directly enters switching fabric exchanges the descending ply-yarn drill of arrival purpose, and this exchanges data scheduling mode is called the Push scheduling mode; Send instruction by the descending ply-yarn drill of purpose and control in the up ply-yarn drill data and whether can enter switching fabric and exchange, this exchanges data scheduling mode is called the Pull scheduling mode.
In switching network, data message has been divided into littler transmission unit (cell) before entering switching fabric, and cell needs to be combined again after arriving descending ply-yarn drill through switching fabric, and according to certain regularly arranged order.
In the switching network structural design, every grade of chip all has the limit value of memory capacity, take as the buffer memory of fruit chip and to have surpassed certain threshold values, just need make this chip no longer receive corresponding cell by certain mode, to avoid too much buffer memory to take or buffer memory overflows, the Flow Control Design in Here it is the switching network.
In the prior art, the flow control of switching network is to adopt the method for back-pressure step by step to realize: buffer memory takies the chip that reaches threshold values and sends the back-pressure signal to higher level's chip direct-connected with it, makes that higher level chip direct-connected with it stops to send data to this chip.
With switching fabric shown in Figure 1 is example, and this switching fabric comprises three stage switching unit SE1, SE2 and SE3 and uplink card (iFIC) and downlink card (eFIC).If the buffer memory of descending ply-yarn drill takies above threshold values, just send the back-pressure signal, so that SE3 stops to send data to descending chip to higher level's chip SE3 direct-connected with it; Similarly, if the buffer memory of SE3 takies above threshold values, SE3 sends the back-pressure signal just as SE2; If the buffer memory of SE2 takies above threshold values, SE2 sends the back-pressure signal just as SE1; If the buffer memory of SE1 takies above threshold values, SE1 sends the back-pressure signal just as up ply-yarn drill, so that up ply-yarn drill stops to send data.
In the switching network structure, it is a very part and parcel that the reorganization of cell of data is reordered.Reorder efficiently is the assurance of switching fabric outlet bandwidth.If reordering, the reorganization of system data cell carries out at descending ply-yarn drill fully, promptly having only descending chip that the cell of data of receiving is recombinated reorders, and crosspoint SE1, SE2 and SE3 are just in the transmission of carrying out cell of data, the flow control methods of back-pressure will cause deadlock as described below and vicious circle so, step by step.
Why descending chip is exactly because piled up the cell that can't successfully organize bag and ordering in a large number to higher level's chip back-pressure, it is pressing for these cells that lacks, and at this moment it has stoped the arrival of cell to the back-pressure of higher level's chip, so it is not only at the wait cell but also at the refusal cell, thereby has just produced deadlock situation.If descending ply-yarn drill has the time restriction of a wait, just abandon the cell that these successfully do not organize bag and ordering after promptly waiting for a period of time, so will inevitably overtime packet loss, after overtime packet loss, descending ply-yarn drill can be cancelled the back-pressure to higher level's chip, so a pile of can getting off again can't successfully be organized the cell (because some relevant cell has been abandoned by overtime) of bag and ordering, has so just caused vicious circle.
Summary of the invention
The embodiment of the invention provides a kind of switching network flow control methods and device.
A kind of switching network flow control methods, this switching network comprises switching node, and this switching node comprises up ply-yarn drill, descending ply-yarn drill and the switching fabric that connects up ply-yarn drill and descending ply-yarn drill, and this method comprises:
Switch network data is received in the downlink clamping;
The buffer memory of judging received data takies whether surpass the Xiamen limit value;
When the buffer memory of received data took above the Xiamen limit value, the transmitted traffic control signal was given up ply-yarn drill, and this flow control signal is used to make up ply-yarn drill to reduce data transmission rate.
A kind of ply-yarn drill as output, comprises receiving element, judging unit and processing unit;
Receiving element is used to receive switch network data;
Judging unit is used to judge the buffer memory of receiving element received data takies whether surpass the Xiamen limit value;
Processing unit is used for when the buffer memory of received data takies above the Xiamen limit value, and the transmitted traffic control signal is given the input ply-yarn drill, and this flow control signal is used to make the input ply-yarn drill to reduce data transmission rate.
A kind of router, comprise up ply-yarn drill, descending ply-yarn drill and the switching fabric that connects up ply-yarn drill and descending ply-yarn drill, this descending ply-yarn drill, be used to receive switch network data, the buffer memory of judging received data takies whether surpass the Xiamen limit value, if then the transmitted traffic control signal is given up ply-yarn drill, this flow control signal is used to make up ply-yarn drill to reduce data transmission rate.
The embodiment of the invention adopts when the buffer memory of descending ply-yarn drill takies above the Xiamen limit value, adjust the technical scheme of the data transmission rate of up ply-yarn drill, the buffer memory of descending ply-yarn drill can be taken and be controlled at below the certain level, thus avoid descending ply-yarn drill upwards the level chip send a deadlock and the vicious circle problem that the back-pressure signal is caused.
Description of drawings
Fig. 1 is the switching fabric schematic diagram of prior art;
Fig. 2 is the flow chart of switching network flow control methods provided by the invention;
Fig. 3 is the switching fabric schematic diagram of the embodiment of the invention one;
Fig. 4 is the switching fabric schematic diagram of the embodiment of the invention two;
Fig. 5 is the logical construction schematic diagram of ply-yarn drill provided by the invention;
Fig. 6 is the logical construction schematic diagram of router provided by the invention.
Embodiment
The embodiment of the invention provides a kind of switching network flow control methods, when employing takies above the Xiamen limit value at the buffer memory of descending ply-yarn drill, adjust the technical scheme of the data transmission rate of up ply-yarn drill, the buffer memory of descending ply-yarn drill can be taken as far as possible and be controlled at below the certain level, thus avoid descending ply-yarn drill upwards the level chip send a deadlock and the vicious circle problem that the back-pressure signal is caused.The embodiment of the invention also provides corresponding device thereof.Below be elaborated respectively.
The switching network flow control methods that the embodiment of the invention provides is applied to comprise in the switching network of switching node, is specially adapted in the switching network that switching node is a router.This switching node comprises up ply-yarn drill, descending ply-yarn drill and the switching fabric that connects up ply-yarn drill and downlink, this switching fabric can be single-stage, back-to-back or multistage.Please refer to Fig. 1, the single-stage switching fabric only comprises the one-level exchange chip, for example only comprises SE2, and does not comprise SE1 and SE3; Switching fabric comprises the two-stage exchange chip back-to-back, for example comprises SE1 and SE3, and does not comprise SE2; Multilevel interchange frame comprises exchange chip more than three grades, for example comprises SE1, SE2 and SE3.
As shown in Figure 2, the switching network flow control methods that provides of the embodiment of the invention comprises:
101, switch network data is received in the downlink clamping.
The switch network data that up ply-yarn drill sends arrives descending ply-yarn drill through switching fabric, and the cell of data that descending ply-yarn drill will successfully be organized bag and ordering forwards, and will successfully not organize the cell of data buffer memory of bag.
102, descending ply-yarn drill is judged that the buffer memory of received data takies and whether is surpassed the Xiamen limit value.
This Xiamen limit value can set in advance according to the cache size of descending ply-yarn drill.
103, when the buffer memory of received data takies above the Xiamen limit value, descending ply-yarn drill transmitted traffic control signal is given up ply-yarn drill, and this flow control signal is used to make up ply-yarn drill to reduce data transmission rate.
It should be noted that flow control signal sends to up ply-yarn drill by descending ply-yarn drill, without the chips at different levels of switching fabric, the chips at different levels to switching fabric do not exert an influence.
As preferred version, this method also can comprise:
104, descending ply-yarn drill is judged that the buffer memory of received data takies and whether is surpassed upper threshold.
This upper threshold also is that the cache size according to descending ply-yarn drill sets in advance, and in the present embodiment, this upper threshold is set greater than the Xiamen limit value.
105, when the buffer memory of received data takies above upper threshold, descending ply-yarn drill send the back-pressure signal to higher level's exchange chip of downlink card connection, this back-pressure signal is used to make higher level's exchange chip to stop to send data.
It should be noted that, this back-pressure signal by descending ply-yarn drill send to the direct-connected higher level's exchange chip of descending ply-yarn drill after, should can upwards send the back-pressure signal step by step until up ply-yarn drill with the direct-connected exchange chip of descending ply-yarn drill, exchange chips at different levels and up ply-yarn drill stop to send data after receiving the back-pressure signal immediately.
Adopt said method, when the buffer memory of downlink card took above the Xiamen limit value, descending ply-yarn drill transmitted traffic control signal gives up ply-yarn drill so that up ply-yarn drill reduces data transmission rate, under extreme case, data transmission rate can be reduced to zero, promptly up ply-yarn drill stops to send data.Understand easily, though up ply-yarn drill has reduced data transmission rate even has stopped the transmission data, but also have data at this moment in the exchange chips at different levels of switching fabric in transmission, thereby descending ply-yarn drill can receive within a certain period of time that still the data that need finish the group bag, in case avoided in the prior art buffer memory to take surpassing threshold value promptly can not receiving data and deadlock problem that produces and the vicious circle that further causes again.
In the preferred version of said method, also comprised when the buffer memory of descending ply-yarn drill takies above upper threshold, sent the back-pressure signal to higher level's exchange chip, and upwards send the back-pressure signal step by step, thereby stop transfer of data, to avoid too much buffer memory to take or buffer memory overflows.
Embodiment one, a kind of switching network flow control methods
In the present embodiment, the data of transmitting in the switching network are unicast data.Unicast data adopts the Pull scheduling mode, i.e. the data of up ply-yarn drill transmission is to be controlled by the flow control signal of descending ply-yarn drill.This flow control signal specifically comprises: send amount control signal (Credit) or arbitration result control signal (Grant).
Referring to shown in Figure 3, switching fabric comprises three stage switching unit (chip) SE1, SE2 and SE3 and uplink card (iFIC) and downlink card (eFIC).Pull scheduling characteristic based on unicast data, from crosspoint SE3 to SE2, from SE2 to SE1, SE1 is to up ply-yarn drill, adopt the amount control method of antibaric flow step by step of prior art, still, the method after improving is adopted in the flow control that snaps into crosspoint SE3 from downlink.
Descending ply-yarn drill takies the unicast data of buffer memory two kinds of threshold values is set, i.e. Xiamen limit value and upper threshold, and wherein the Xiamen limit value is used to adjust Credit or Grant corresponding to flow control signal; Upper threshold is corresponding to direct back-pressure signal to its higher level's chip.
Descending ply-yarn drill takies situation according to buffer memory and clocklike adjusts flow control signal Credit or Grant, to reach the purpose that makes the corresponding adjustment data transmission rate of up ply-yarn drill.In the time of can taking above the Xiamen limit value at the buffer memory of descending ply-yarn drill, Credit or Grant are reduced to certain level, thereby make the data transmission rate of up ply-yarn drill be reduced to certain level.
As preferred version, the Xiamen limit value can comprise one group of concrete threshold value, and for example three numerical value A, B, the C that increases successively supposes that A, B, C are respectively 65%, 70%, 75% of descending ply-yarn drill cache size.When the buffer memory of downlink card takies when reaching A, just Credit or Grant are reduced to certain level and for example set 60% of numerical value for initial value or certain; When the buffer memory of downlink card takies when reaching B, just Credit or Grant are reduced to certain level and for example set 30% of numerical value for initial value or certain; When the buffer memory of downlink card takies when reaching C, just Credit or Grant are reduced to certain level and for example directly are reduced to zero.Thereby make the data transmission rate of up ply-yarn drill be reduced to certain level successively, for example 60% of initial value, 30%, or even zero.Summarize, in the above-mentioned preferred version, the Xiamen limit value can comprise one group of concrete threshold value, the corresponding concrete flow control signal of each concrete threshold value, and this concrete flow control signal is used to make up ply-yarn drill that data transmission rate is reduced to an occurrence.
Certainly, the buffer memory of descending ply-yarn drill takies under a stable condition still may reach upper threshold, and at this moment, descending ply-yarn drill can send the back-pressure signal to its higher level's exchange chip SE3 immediately, and SE3 receives that the back-pressure signal stops to send unicast data immediately to descending ply-yarn drill.SE3 step by step sends back-pressure signal to SE1, SE1 to the uplink card to SE2, SE2 simultaneously.
The method of present embodiment, by two kinds of threshold values of reasonable disposition, for example Xiamen limit value and upper threshold are set to 70% and 90% of descending ply-yarn drill cache size respectively, the unicast data buffer memory take reach upper threshold before, descending ply-yarn drill be adjusted Credit or Grant, send flow control signal to up ply-yarn drill, make up ply-yarn drill reduce data transmission rate, be controlled at upper threshold below thereby the buffer memory of descending ply-yarn drill is taken as far as possible, and then avoid the generation of deadlock and vicious circle problem.
Embodiment two, a kind of switching network flow control methods
In the embodiment of the invention, the data of transmitting in the switching network are multi-case data.Unicast data adopts the Push scheduling mode, and promptly the data of up ply-yarn drill directly send in the switching fabric.
Referring to shown in Figure 4, switching fabric comprises three stage switching unit (chip) SE1, SE2 and SE3 and uplink card (iFIC) and downlink card (eFIC).Push scheduling characteristic based on multi-case data, from crosspoint SE3 to SE2, from SE2 to SE1, SE1 is to up ply-yarn drill, adopt prior art antibaric flow amount control method step by step, still, the method after improving is adopted in the flow control that snaps into crosspoint SE3 from downlink.
Descending ply-yarn drill takies the multi-case data of buffer memory two kinds of threshold values, i.e. Xiamen limit value and upper threshold is set.As preferably, the message number threshold value of not finishing the group bag can also be set.The cache size of the mean size of message and descending ply-yarn drill is relevant in message number threshold value and the system, can dispose according to actual scene.
When taking, the multi-case data buffer memory of downlink card reached the Xiamen limit value, then descending ply-yarn drill sends back-pressure signal (flow control signal) to all up ply-yarn drills, so that up ply-yarn drill stops to send multi-case data downwards, promptly the data transmission rate of up ply-yarn drill is zero.But this moment, may also have multi-case data in the multistage chip of switching fabric in transmission, thereby downlink is stuck in the certain hour and can also receives multi-case data.
As preferably, in the multi-case data buffer memory of downlink card, not finish the message number of organizing bag and reached message number threshold value, descending ply-yarn drill also sends the back-pressure signal to all up ply-yarn drills, so that up ply-yarn drill stops to send multi-case data downwards.Same, this moment switching fabric multistage chip in may also have multi-case data in transmission, thereby downlink is stuck in the certain hour and can also receives multi-case data.
Certainly, the buffer memory of descending ply-yarn drill takies under a stable condition still may reach upper threshold, and at this moment, descending ply-yarn drill can send the back-pressure signal to its higher level's exchange chip SE3 immediately, and SE3 receives that then the back-pressure signal stops to send multi-case data immediately to descending ply-yarn drill.SE3 step by step sends back-pressure signal to SE1, SE1 to the uplink card to SE2, SE2 simultaneously.
The method of present embodiment, by upper and lower threshold value of reasonable disposition and message number threshold value, for example Xiamen limit value and upper threshold are set to 70% and 90% of descending ply-yarn drill cache size respectively, the multi-case data buffer memory take reach upper threshold before, descending ply-yarn drill has promptly sent the back-pressure signal to up ply-yarn drill, make up ply-yarn drill stop to send multi-case data, be controlled at upper threshold below thereby the buffer memory of descending ply-yarn drill is taken as far as possible, and then avoid the generation of deadlock and vicious circle problem.
Embodiment three, a kind of ply-yarn drill
Present embodiment provides a kind of ply-yarn drill, as output, as shown in Figure 5, comprises receiving element 201, judging unit 202 and processing unit 203.Wherein, receiving element 201 is used to receive switch network data; Judging unit 202 is used to judge the buffer memory of received data takies whether surpass the Xiamen limit value; Processing unit 203 is used for when the buffer memory of received data takies above the Xiamen limit value, and the transmitted traffic control signal is given up ply-yarn drill, and this flow control signal is used to make up ply-yarn drill to reduce data transmission rate.
In a preferred version, judging unit 202 is used to also judge the buffer memory of received data takies whether surpass upper threshold; Processing unit 203 also is used for when the buffer memory of received data takies above upper threshold, sends the back-pressure signal to the higher level's exchange chip that is connected with ply-yarn drill, and this back-pressure signal is used to make higher level's exchange chip to stop to send data.
The ply-yarn drill that the employing present embodiment provides is used for switching network as the descending ply-yarn drill of router with this router, can avoid produce of deadlock and vicious circle problem.
Embodiment four, a kind of router
Present embodiment provides a kind of router, as shown in Figure 6, comprises up ply-yarn drill 301, descending ply-yarn drill 302 and the switching fabric 303 that is connected up ply-yarn drill and descending ply-yarn drill.Wherein descending ply-yarn drill 302, be used to receive switch network data, judge the buffer memory of received data takies whether surpass the Xiamen limit value, if, then the transmitted traffic control signal is given up ply-yarn drill 301, and this flow control signal is used to make up ply-yarn drill to reduce data transmission rate.
In a preferred version, switching fabric 303 comprises the above exchange chip of one-level; Descending ply-yarn drill 302 is used to also judge the buffer memory of received data takies whether surpass upper threshold, if, then sending the back-pressure signal to the higher level's exchange chip that is connected with descending ply-yarn drill 302, this back-pressure signal is used to make higher level's exchange chip to stop to send data.
The router that adopts present embodiment to provide is used for switching network, can avoid produce of deadlock and vicious circle problem.
The switching network flow control methods of the embodiment of the invention adopts when the buffer memory of descending ply-yarn drill takies above the Xiamen limit value, and the transmitted traffic control signal makes it reduce data transmission rate for up ply-yarn drill; When the buffer memory of descending ply-yarn drill takies above upper threshold, send the back-pressure signal and make it stop to send data to higher level's exchange chip; Wherein upper threshold is greater than the technical scheme of Xiamen limit value, the buffer memory of descending ply-yarn drill can be taken as far as possible to be controlled at below the certain level, thereby avoid descending ply-yarn drill upwards the level chip send a deadlock and the vicious circle problem that the back-pressure signal is caused.
For unicast traffic, descending ply-yarn drill can take situation according to buffer memory and at first adjust flow control signal Credit or Grant, reduces or stop the data of up ply-yarn drill transmission, eliminates the upwards back-pressure of level chip of descending ply-yarn drill as far as possible.
For multicast traffic, descending ply-yarn drill takies situation according to buffer memory and the current message amount of not finishing the group bag makes up ply-yarn drill stop to send data, eliminates the upwards back-pressure of level chip of descending ply-yarn drill as far as possible.
The switching network flow control methods of the embodiment of the invention on traditional counter pressure flow controlling method basis, by setting up threshold value, is implemented to dam on up ply-yarn drill, avoids the congested of swap data unit, has guaranteed time delay end to end.
More than a kind of switching network flow control methods, ply-yarn drill and router that the embodiment of the invention provided are described in detail, used specific case herein principle of the present invention and execution mode are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (11)
1. switching network flow control methods, this switching network comprises switching node, described switching node comprises up ply-yarn drill, descending ply-yarn drill and the switching fabric that connects described up ply-yarn drill and descending ply-yarn drill, it is characterized in that this method comprises:
Switch network data is received in described downlink clamping;
The buffer memory of judging received data takies whether surpass the Xiamen limit value;
When the buffer memory of received data took above described Xiamen limit value, the transmitted traffic control signal was given described up ply-yarn drill, and described flow control signal is used to make described up ply-yarn drill to reduce data transmission rate.
2. method according to claim 1 is characterized in that, described switching fabric comprises the above exchange chip of one-level, and described method also comprises:
The buffer memory of judging received data takies whether surpass upper threshold, and described upper threshold is greater than described Xiamen limit value;
Take when surpassing described upper threshold at the buffer memory of received data, send the back-pressure signal to higher level's exchange chip of described downlink card connection, described back-pressure signal is used to make described higher level's exchange chip to stop to send data.
3. method according to claim 1 and 2 is characterized in that: described switch network data is a unicast data, and described flow control signal comprises: send amount control signal and arbitration result control signal.
4. method according to claim 3, it is characterized in that, described Xiamen limit value comprises one group of concrete threshold value, the corresponding concrete flow control signal of each concrete threshold value, and described concrete flow control signal is used to make described up ply-yarn drill that data transmission rate is reduced to an occurrence.
5. method according to claim 1 and 2 is characterized in that: described switch network data is a multi-case data, and described flow control signal is specially the back-pressure signal.
6. method according to claim 5 is characterized in that, described up ply-yarn drill reduces data transmission rate and is specially: described up ply-yarn drill is reduced to zero with data transmission rate.
7. method according to claim 5 is characterized in that, this method also comprises:
Judge whether the message number of not finishing the group bag in the buffer memory surpasses message number threshold value, if, then sending the back-pressure signal to described up ply-yarn drill, described back-pressure signal is used to make described up ply-yarn drill to stop to send data.
8. a ply-yarn drill as output, is characterized in that: comprise receiving element, judging unit and processing unit;
Described receiving element is used to receive switch network data;
Described judging unit is used to judge the buffer memory of described receiving element received data takies whether surpass the Xiamen limit value;
Described processing unit is used for when the buffer memory of received data takies above described Xiamen limit value, and the transmitted traffic control signal is given the input ply-yarn drill, and described flow control signal is used to make described input ply-yarn drill to reduce data transmission rate.
9. ply-yarn drill according to claim 8 is characterized in that:
Described judging unit is used to also judge the buffer memory of received data takies whether surpass upper threshold;
Described processing unit also is used for sending the back-pressure signal to the higher level's exchange chip that is connected with described ply-yarn drill when the buffer memory of received data takies above described upper threshold, and described back-pressure signal is used to make described higher level's exchange chip to stop to send data.
10. a router comprises up ply-yarn drill, descending ply-yarn drill and the switching fabric that connects described up ply-yarn drill and descending ply-yarn drill, it is characterized in that:
Described descending ply-yarn drill is used to receive switch network data, judges the buffer memory of received data takies whether surpass the Xiamen limit value, if then the transmitted traffic control signal is given up ply-yarn drill, described flow control signal is used to make described up ply-yarn drill to reduce data transmission rate.
11. router according to claim 10 is characterized in that:
Described switching fabric comprises the above exchange chip of one-level;
Described descending ply-yarn drill, be used to also judge the buffer memory of received data takies whether surpass upper threshold, if, then send the back-pressure signal to higher level's exchange chip of described downlink card connection, described back-pressure signal is used to make described higher level's exchange chip to stop to send data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010105273825A CN101984594A (en) | 2010-10-25 | 2010-10-25 | Switching network flow control method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010105273825A CN101984594A (en) | 2010-10-25 | 2010-10-25 | Switching network flow control method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101984594A true CN101984594A (en) | 2011-03-09 |
Family
ID=43641765
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010105273825A Pending CN101984594A (en) | 2010-10-25 | 2010-10-25 | Switching network flow control method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101984594A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104052676A (en) * | 2014-06-13 | 2014-09-17 | 华为技术有限公司 | Transmitting channel and data processing method thereof |
| WO2016000205A1 (en) * | 2014-07-01 | 2016-01-07 | 华为技术有限公司 | Data transmission control method, passive optical network equipment and device, and passive optical network |
| CN105610820A (en) * | 2015-12-28 | 2016-05-25 | 中国电子科技集团公司第五十四研究所 | Multipath transport control protocol (MPTCP) based congestion control method and apparatus |
| WO2016082365A1 (en) * | 2014-11-25 | 2016-06-02 | 中兴通讯股份有限公司 | Traffic adjusting method and device |
| CN107770090A (en) * | 2017-10-20 | 2018-03-06 | 深圳市楠菲微电子有限公司 | Method and apparatus for controlling register in streamline |
| CN108073445A (en) * | 2016-11-18 | 2018-05-25 | 腾讯科技（深圳）有限公司 | The back pressure processing method and system calculated based on distributed stream |
| CN108153689A (en) * | 2016-12-06 | 2018-06-12 | 比亚迪股份有限公司 | The method and apparatus of poll arbitration |
| CN109240829A (en) * | 2018-08-29 | 2019-01-18 | 盛科网络（苏州）有限公司 | For exchange chip application, manage exclusive resource method and device |
| CN110022261A (en) * | 2019-05-20 | 2019-07-16 | 北京邮电大学 | Multi-path transmission method and apparatus based on SCTP-CMT transport protocol |
| CN111526097A (en) * | 2020-07-03 | 2020-08-11 | 新华三半导体技术有限公司 | Message scheduling method, device and network chip |
| CN113206794A (en) * | 2021-03-31 | 2021-08-03 | 新华三信息安全技术有限公司 | Forwarding speed limiting method and device |
| CN114731294A (en) * | 2020-01-21 | 2022-07-08 | 华为技术有限公司 | Exchange network chip and exchange equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1627745A (en) * | 2003-12-11 | 2005-06-15 | 上海贝尔阿尔卡特股份有限公司 | Method for controlling data flux |
| US20070097864A1 (en) * | 2005-11-01 | 2007-05-03 | Cisco Technology, Inc. | Data communication flow control |
| CN101621833A (en) * | 2009-08-13 | 2010-01-06 | 中兴通讯股份有限公司 | Message flux control method and base station controller |
-
2010
- 2010-10-25 CN CN2010105273825A patent/CN101984594A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1627745A (en) * | 2003-12-11 | 2005-06-15 | 上海贝尔阿尔卡特股份有限公司 | Method for controlling data flux |
| US20070097864A1 (en) * | 2005-11-01 | 2007-05-03 | Cisco Technology, Inc. | Data communication flow control |
| CN101621833A (en) * | 2009-08-13 | 2010-01-06 | 中兴通讯股份有限公司 | Message flux control method and base station controller |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104052676A (en) * | 2014-06-13 | 2014-09-17 | 华为技术有限公司 | Transmitting channel and data processing method thereof |
| WO2015188770A1 (en) * | 2014-06-13 | 2015-12-17 | 华为技术有限公司 | Sending path and data processing method for sending path |
| US9813349B2 (en) | 2014-06-13 | 2017-11-07 | Huawei Technologies Co., Ltd. | Transmission path and data processing method for transmission path |
| CN104052676B (en) * | 2014-06-13 | 2017-12-05 | 华为技术有限公司 | A kind of data processing method of transmission path device and transmission path |
| WO2016000205A1 (en) * | 2014-07-01 | 2016-01-07 | 华为技术有限公司 | Data transmission control method, passive optical network equipment and device, and passive optical network |
| US10491323B2 (en) | 2014-07-01 | 2019-11-26 | Huawei Technologies Co., Ltd. | Data transmission control method, passive optical network equipment, apparatus, and passive optical network |
| WO2016082365A1 (en) * | 2014-11-25 | 2016-06-02 | 中兴通讯股份有限公司 | Traffic adjusting method and device |
| CN105610820A (en) * | 2015-12-28 | 2016-05-25 | 中国电子科技集团公司第五十四研究所 | Multipath transport control protocol (MPTCP) based congestion control method and apparatus |
| CN105610820B (en) * | 2015-12-28 | 2019-03-22 | 中国电子科技集团公司第五十四研究所 | A kind of jamming control method and device based on MPTCP |
| CN108073445A (en) * | 2016-11-18 | 2018-05-25 | 腾讯科技（深圳）有限公司 | The back pressure processing method and system calculated based on distributed stream |
| CN108073445B (en) * | 2016-11-18 | 2021-10-22 | 腾讯科技（深圳）有限公司 | Backpressure processing method and system based on distributed flow calculation |
| CN108153689A (en) * | 2016-12-06 | 2018-06-12 | 比亚迪股份有限公司 | The method and apparatus of poll arbitration |
| CN108153689B (en) * | 2016-12-06 | 2020-04-24 | 比亚迪股份有限公司 | Method and apparatus for polling arbitration |
| CN107770090B (en) * | 2017-10-20 | 2020-05-01 | 深圳市楠菲微电子有限公司 | Method and apparatus for controlling registers in a pipeline |
| CN107770090A (en) * | 2017-10-20 | 2018-03-06 | 深圳市楠菲微电子有限公司 | Method and apparatus for controlling register in streamline |
| CN109240829A (en) * | 2018-08-29 | 2019-01-18 | 盛科网络（苏州）有限公司 | For exchange chip application, manage exclusive resource method and device |
| CN109240829B (en) * | 2018-08-29 | 2021-02-02 | 盛科网络（苏州）有限公司 | Method and device for applying for exchanging chip and managing exclusive resource |
| CN110022261A (en) * | 2019-05-20 | 2019-07-16 | 北京邮电大学 | Multi-path transmission method and apparatus based on SCTP-CMT transport protocol |
| CN114731294A (en) * | 2020-01-21 | 2022-07-08 | 华为技术有限公司 | Exchange network chip and exchange equipment |
| CN111526097A (en) * | 2020-07-03 | 2020-08-11 | 新华三半导体技术有限公司 | Message scheduling method, device and network chip |
| CN113206794A (en) * | 2021-03-31 | 2021-08-03 | 新华三信息安全技术有限公司 | Forwarding speed limiting method and device |
| CN113206794B (en) * | 2021-03-31 | 2022-05-27 | 新华三信息安全技术有限公司 | Forwarding speed limiting method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101984594A (en) | Switching network flow control method and device | |
| CN101536413B (en) | Queue aware flow control | |
| US7298756B1 (en) | System and method for programming cell packet headers | |
| US6496516B1 (en) | Ring interface and ring network bus flow control system | |
| CN101631081B (en) | Multistage switched network | |
| US8174987B2 (en) | Method and apparatus for implementing output queue-based flow control | |
| CN105337883A (en) | Multi-business supporting network switching device and implementation method therefor | |
| CN101621833B (en) | Message flux control method and base station controller | |
| CN102726009B (en) | Data flow exchange control method and relative device | |
| EP1891778B1 (en) | Electronic device and method of communication resource allocation. | |
| EP2093945B1 (en) | A method for realizing back pressure of masses of ports and the device thereof | |
| CN102811176B (en) | A kind of data flow control method and device | |
| CN102111327B (en) | Method and system for cell dispatching | |
| CN101222436A (en) | Graded data transmission method | |
| CN104717159A (en) | Scheduling method based on store-and-forward switching structure | |
| CN105530207A (en) | Switch port arbitration circuit and method in support of multi-type data frames | |
| CN100417136C (en) | Down queue fast back pressure transmitting based on three-stage exchange network | |
| CN103746922A (en) | Method and device for forwarding multicast messages | |
| CN101478486B (en) | Method, equipment and system for switch network data scheduling | |
| EP2477366B1 (en) | Data transmission method, apparatus and system | |
| CN101616343A (en) | A kind of auto-negotiation method for service types and convergence device thereof | |
| CN102957625A (en) | System and method for implementing end-to-end hierarchical quality of service | |
| CN109391559A (en) | The network equipment | |
| CN101854402B (en) | Interface conversion device and stream control implementing method | |
| CN110430146B (en) | Cell recombination method based on CrossBar switch and switch structure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20110309 |