CN1536816A - Flow shaping method in data transmission process - Google Patents
Flow shaping method in data transmission process Download PDFInfo
- Publication number
- CN1536816A CN1536816A CNA031092004A CN03109200A CN1536816A CN 1536816 A CN1536816 A CN 1536816A CN A031092004 A CNA031092004 A CN A031092004A CN 03109200 A CN03109200 A CN 03109200A CN 1536816 A CN1536816 A CN 1536816A
- Authority
- CN
- China
- Prior art keywords
- message
- unit
- sent
- transmit queue
- storage organization
- 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.)
- Granted
Links
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a flow shaping method in data transmission process, the method includes following steps: a) creates a storage structure for storing waiting message, each unit in the transmitting queue is correspondent to a storage structure; b) determines the unit which can buffer waiting message in the transmission queue, stores the message into the storage structure correspondent to the buffer unit; c) dispatches the preset message transmission in the storage structure according to the dispatching tactics. The invention has several storage structures and sets different transmission times in order to guarantee the transmission is correct, it can reduce the length of transmission queue greatly, reduces the occupation of inner memory.
Description
Technical field
The present invention relates to a kind of method of transfer of data, relate to the method for flow control in a kind of data transmission procedure more specifically.
Background technology
Now, the data communication technology development is very fast, is constantly all transmitting mass data in communication network, and therefore, handling the transmission of data in communication network well is the important means that improves network resource utilization and data transmission efficiency.Along with the transmission of mass data in network, tend to occur such situation, because network capacity is limited, when bursty traffic, a lot of data often have little time to send and just are dropped; Also, need send control, for example require to send data according to the specific moment, speed, sequential scheduling to the data of transmitting in the network sometimes in order to satisfy the customized network service quality of user, or the like.When above-mentioned situation occurring, common solution be the network bursty data or according to service quality require the transmission of control data the time, with the metadata cache that has little time to send, network is idle or send in the suitable moment again by the time.The emphasis of traffic shaping is exactly that the moment that metadata cache and control data send is combined, and is no more than the size of agreement to guarantee streaming rate.
Fig. 1 is the dispatching method schematic diagram of prior art traffic shaping, this method relates to shaping unit, scheduling part and a transmit queue, shaping unit is used for determining the moment of message transmission, and message inserted units corresponding in the transmit queue, and the scheduling part is responsible in the moment of regulation the message in the transmit queue being sent, each unit in the transmit queue is the buffer area of a fixed size, can only store a message, during transmission, the scheduling part sends the message in the transmit queue successively, and the time that guarantees a unit of scheduling be constant, be made as T0.A certain unit is for empty in the transmit queue as if being checked through, and expression does not have message to need to send, and waits for that then T0 just dispatches the message transmission of next unit after the time.
As shown in Figure 1, dispatch part successively with message in the transmit queue 1,2,3,4,5 ... send.Suppose that queue length is N, after the scheduling part sends N message, go back to the transmission that the formation stem begins next round again.
Traffic shaping schematic diagram among the ATM as shown in Figure 2 (asynchronous transfer mode), realize traffic shaping for the common method of utilizing constant message transmitting spped rate, when shaping unit attempts that message is put in a certain unit in transmit queue, if the existing message in this unit exists, then begin to find the unit of a sky from this unit backward till.If the unit of search has arrived last unit of formation, then search for again, but can not surpass the unit of dispatching the current transmission of part from the beginning of formation.In Fig. 2, M1 is the unit of current transmission, and M2 is the unit that shaping unit attempts message is put into formation, and the scope that shaping unit is searched for non-dummy cell can only be M2, (M2+1) ..., N, 1,2, in (M1-1) one.
With traffic shaping among the ATM shown in Figure 2 is that example is analyzed.In above-mentioned technology, the ability of traffic shaping is subjected to the direct restriction of queue length, a kind of situation below considering, as shown in Figure 3, at a time, data source X produces bigger message burst, and burst length is N, also is the length of transmit queue, be assumed to even number, in addition, do not have other data flow.M1 is the current transmitting element of formation (2<M1<N) among Fig. 3, suppose and to carry out the shaping of CBR (constant cell rate) to data stream, and SCR (continue cell rate) just has been 1/2 of transmission rate, in promptly per two dispatching message cycle 2T, the message of data source X can be sent.Shaping unit is the message X (0) of this data flow, X (1), and X (2) ..., X (N/2-2), X (N/2-1) inserts the unit M1 in the transmit queue successively, (M1+2), (M1+4) ..., (M1-4), (M1-2) in.When shaping unit attempted that (N/2) individual message put into transmit queue, the unit that can find to put into was also at M1, but X (0) is in this unit.So, must be with message X (N/2) and subsequent packet X (N/2+1) thereof ..., X (N) abandons.In this case, data source X hereto, length is the burst length that the transmit queue of N can only be tolerated (N/2-1) individual message at most.
Therefore, in above-mentioned technology, the length of transmit queue has determined the Maximum Burst Size that can tolerate fully.And improve simple method of the anti-burst capability of traffic shaping is exactly to strengthen the length of transmit queue, and the length of formation is long more, and the ability of the anti-burst of traffic shaping is just strong more.But along with the increase of burst length, it is very huge that the length of transmit queue may become.
On the other hand, bigger data burst is occurrent, and what the probability that occurs in the long time may be very is low.When strengthening transmit queue length, the decline of internal memory utilization ratio and the increase of cost have also been caused.
Summary of the invention
The purpose of this invention is to provide a kind of flow shaping method in data transmission process, be subjected to the restriction of transmit queue length and the lower problem of the utilization ratio of internal memory to solve traffic shaping ability that above-mentioned prior art exists.
The objective of the invention is to realize by following method, a kind of flow shaping method in data transmission process, this method may further comprise the steps:
A) set up storage structure for storing waiting message, the corresponding described storage organization in each unit in the transmit queue;
B) but determine the unit of buffer memory message to be sent in the transmit queue, then message to be sent is deposited in the storage organization of described cacheable unit correspondence;
C) send according to the predetermined message of storing in each unit corresponding stored structure in the scheduling strategy scheduling transmit queue of setting.
The caching of the concrete described message to be sent of step b) is realized according to following step:
(11) but in the unit of described transmit queue search buffer memory message to be sent;
(12) do not have in the storage organization as the unit correspondence that searches message to be sent or not with the identical message of transmission wheel sub-value of the message of desiring to deposit in to be sent, then this message to be sent is deposited in the storage organization of described unit correspondence in the end caching;
As existing and the identical message of transmission wheel sub-value of desiring to deposit in message to be sent in the storage organization of the unit correspondence that searches, then repeat above-mentioned steps (11), (12) but continue to search for buffer unit to deposit this message to be sent in.
Optimize, the described predetermined message of step c) is that transmission wheel sub-value is 0 message.
The present invention compared with prior art has following advantage:
1. the ability of the anti-burst of traffic shaping is not subjected to the restriction of transmit queue length, set up the storage organization of each unit in the corresponding transmit queue among the present invention, because described storage organization can be stored a plurality of messages, be that a plurality of messages are represented in each unit among the present invention, compared to existing technology, the length of transmit queue of the present invention has significantly reduced, when the buffer memory bursty data because the transmission round (roundcount) that message is set is different value, when dispatching each unit message and sending be that 0 message sends with roundcount value in this unit, all the other messages can guarantee that according to sending the round transmission afterwards of wheel sub-value message correctly sends;
2. a further advantage of the invention is the utilization ratio that has improved internal memory, the bursty data odds is seldom, owing to reduced the length of transmit queue, the efficient of the utilization of internal memory is also corresponding have been improved among the present invention.
Description of drawings
Fig. 1 is the dispatching method schematic diagram of prior art traffic shaping;
Fig. 2 is the schematic diagram of traffic shaping in the atm network;
Fig. 3 is the shaping schematic diagram when having the flow shaping method bursty data now;
Fig. 4 is the schematic diagram of the corresponding storage organization in each unit of transmit queue of the present invention;
Fig. 5 is the structure chart that the present invention will send wheel secondary data and message encapsulation;
Fig. 6 is the flow chart of buffer memory of the present invention message to be sent;
Fig. 7 is the flow chart of the specific embodiment of the invention;
Embodiment
In data transmission procedure, traffic shaping is the method for the normal control data stream of using, and still produces the situation of bigger burst in the present embodiment with data source X recited above, in conjunction with attached
Figure is described in detail the method for traffic shaping of the present invention.
For the technical staff is more convenient for understanding method of the present invention, be that example describes still below with ATM (asynchronous transfer mode) network, we have a look the method for existing traffic shaping again, as shown in Figure 3, data source X produces bigger message burst, burst length is N, queue length also is N, data stream is carried out the shaping of CBR (constant cell rate), and SCR (continuing the cell rate) just has been 1/2 of a transmission rate, be in per two dispatching message cycle 2T, the message of data source X can be sent.Shaping unit is the message X (0) of this data flow, X (1), and X (2) ..., X (N/2-2), X (N/2-1) inserts the unit M1 in the transmit queue successively, (M1+2), (M1+4) ..., (M1-4), (M1-2) in.When shaping unit attempted that (N/2) individual message put into transmit queue, the unit that can find to put into was also at M1, but X (0) is in this unit.So, must be with message X (N/2) and subsequent packet X (N/2+1) thereof ..., X (N) abandons.In this case, data source X hereto, length is the burst length that the transmit queue of N can only be tolerated (N/2-1) individual message at most.
Below we analyze the reason that causes the problems referred to above earlier, be not difficult to find, the most direct reason that causes the problems referred to above is but that the unit of buffer memory message to be sent in the transmit queue can only message of buffer memory, the ability that does not have a plurality of messages of buffer memory, aforesaid, when shaping unit attempted that (N/2) individual message put into transmit queue, the unit that can find to put into was also at M1, but X (0) is in this unit.Like this, have to message X (N/2) and subsequent packet X (N/2+1) thereof ..., X (N) abandons.This problem of simple analysis, we can find two kinds of possible reasons:
1) each unit of transmit queue ability of not storing a plurality of messages like this, obviously only abandons when new message;
2) even but there is the ability of a plurality of buffer memorys message to be sent each unit of formation, but after depositing two different messages in, because the asynchronism(-nization) that each message sends, when scheduling sends, the scheduling part can not distinguish send which message, can only abandon message afterwards like this.
Address this problem, the length of the increasing transmit queue that proposes as prior art is a kind of simple way, but as we analyze, the shortcoming of doing like this is exactly that queue length is long, and the bursty data occurrence probability is not very big, this has just caused the waste of internal memory, and just cure the symptoms, not the disease, therefore, we consider under the condition that does not increase transmit queue length, a kind of storage organization that can a plurality of messages to be sent of buffer memory is provided, be that the present invention proposes a kind of improved method, set up a kind of new message storage organization to be sent, can be when not increasing transmit queue length, originally packet buffer a plurality of to be sent that can not buffer memory is got up, and send in the suitable moment.Please refer to shown in Figure 4, specifically, the message storage organization to be sent that the present invention proposes is each unit of corresponding transmit queue, also be that each unit of transmit queue all can find its storage organization of correspondence, can realize corresponding between the corresponding storage organization of each unit and its by the pointer that the storage organization corresponding with it pointed in storage in each unit of transmit queue.Also can realize corresponding between the corresponding storage organization of each unit and its by the mark of the storage storage organization corresponding in each unit of transmit queue with it, such as, can carry out mark to each unit corresponding stored structure, in the unit of transmit queue, store this mark then, when the scheduling message sends, can find the storage organization of this unit correspondence according to this mark.Correspondence with each unit of transmit queue and its storage organization is by store how to dispatch the message transmission when pointer that points to the storage organization corresponding with it is example explanation shaping in each unit of transmit queue below, after shaping unit deposits message to be sent in the storage organization of this unit correspondence, the pointer of this unit storage structure is pointed in storage in this unit, during the scheduling message, when being dispatched to this unit, the scheduling part at first reads the pointer of storing in this unit, further found the message storage organization of this message to be sent of buffer memory by this pointer, being read by this message storage organization needs the predetermined message that sends and dispatches away transmission again.
After setting up storage structure for storing waiting message, when bursty data, determine to deposit this unit of message to be sent in transmit queue in, and this message to be sent is stored in the storage organization of this unit correspondence, during transmission, according to certain scheduling strategy the predetermined message of storing in this unit being gone out team sends, this scheduling strategy may be to send according to the queue unit order, also may be other possible array dispatching methods, owing to do not relate to flesh and blood of the present invention, and be known technology, no longer describe in detail here.
The storage organization of the message to be sent of above-mentioned proposition, it can be a chained list, also can be an array, even can be a fifo queue that the characteristics of these storage organizations are to store a plurality of messages, and need not strengthen transmit queue length, for example, the traffic shaping in the atm network of stating of more than chatting face to face is an example, when shaping unit attempts that (N/2) individual message put into transmit queue, the unit that can find to put into is at M1, and X (0) is in this unit.Because the message storage organization to be sent of the present invention's structure can be stored a plurality of messages, for example, the storage organization of unit M1 correspondence is assumed to be one-dimension array, then deposit X (0) first array location of this array in after, when X (N/2) arrives, also can deposit in this array, but what at this moment deposit in is next array location, obviously, its subsequent packet X (N/2+1),, X (N) also can deposit in according to identical principle in the transmit queue corresponding units, and can not be dropped.
In the traffic shaping, the delivery time of message is to carry out according to certain algorithm, as GCRA (generic cell rate algorithm) algorithm etc., after receiving message to be sent, need to determine to deposit in the transmit queue unit of message to be sent, in the prior art, because the restriction of transmit queue length, surpass certain bursty data amount, some messages just have to be dropped, the message storage organization of being set up among the present invention then can these messages of buffer memory, and in the suitable moment, these messages are sent, certainly, here for the correct order of transmission of distinguishing message, we can be according to the length of transmit queue after the transmitting time of determining message, and the present time, draw the transmission wheel sub-value of message to be sent, for example, the object time t1 definite according to shaping algorithm, current time t0, the transmit queue length N is fixed owing to dispatch the time of each unit in this transmit queue, is made as T0, the time that queue scheduling one is taken turns is T, then get the transmission wheel sub-value of the integer part of (t1-t0)/T as message, the transmission of message wheel sub-value is 0 so as described, represents that this message sent in the first round, be that 1 this message of expression is taken turns transmission second
After having determined the transmission wheel sub-value of message to be sent, should send wheel secondary data and message to be sent and be made as one, multiple mode can be arranged, a kind of mode as shown in Figure 5, in the heading of message to be sent, be encapsulated into this data, the i.e. transmission of this message wheel sub-value, read the data in the heading during transmission, just can obtain the transmission wheel sub-value of this message, perhaps when message to be sent is written into internal memory, in this transmission wheel sub-value of place, the address mark of depositing this message, also obtained the transmission wheel sub-value of this message in the time of the address that obtains this message during transmission, briefly, described transmission round with message to be sent and this message is made as one and has set up the one-to-one relationship that sends round and message to be sent, actually can also enumerate more example, here be representative only with above-mentioned example, and be not limited to above-mentioned execution mode, further specify method of the present invention below, the above-mentioned transmission wheel sub-value of having determined message, and constructed the message storage organization that to store many messages, further, also needed message to be sent is deposited in the cacheable unit of transmit queue, the flow chart of buffer memory message to be sent as shown in Figure 6 comprises the steps:
Step (11) but in transmit queue the unit of search buffer memory message to be sent, as above-mentioned, such unit is corresponding with a storage organization, be example also with above-mentioned atm network, if the unit p0 of current queue scheduling, the length of transmit queue is N and finishes a period T of taking turns queue scheduling and can determine this is treated that buffer memory sends the object element p1 that message will deposit formation in, specifically determines according to following method:
The remainder that note (t1-t0)/T obtains is r, and then p1 gets the (integer part of (r/T0)+p0); If the p1>N that obtains like this, then p1=p1-N down looks for buffer unit but begin search from described object element p1 like this;
Step (12) but as searching buffer unit, and there is not message to be sent in the storage organization of this unit correspondence, then execution in step (14) deposits message to be sent in the storage organization of this unit correspondence, in fact such situation is seldom, as the problem to be solved in the present invention, often have message to be sent in the storage organization of this unit correspondence, in the prior art, the way of taking is a dropping packets, and among the present invention based on abovementioned steps, whether the existing message that execution in step (13) is at first checked in this storage organization is identical with the transmission wheel sub-value of desiring to deposit in message, as difference, then execution in step (14) deposits message to be sent in this unit corresponding stored structure in, finish caching then, can guarantee that like this it is different that the message of storing in the storage organization of this unit correspondence sends the wheel sub-value, take turns the sub-value transmission and can't cause confusion according to sending during transmission, but as in the storage organization of the buffer unit correspondence that searches existing message and the transmission of desiring to deposit in message to take turns sub-value identical, then repeat above-mentioned steps (11), (12), (13) but continue the search buffer unit to deposit this message to be sent in.
Further describe method of the present invention below, for ease of understanding, still atm network is carried out traffic shaping is example with above-mentioned, with reference to shown in Figure 7
Step 701, set up a transmit queue earlier, length is made as N, and set up the storage organization of each unit of corresponding transmit queue, here storage organization is example with the chained list, then according to shaping algorithm, determines to send the transmitting time of message, and further determine to send the transmission wheel sub-value of message, should send the round Value Data then and message to be sent is made as one;
Step 702, but in transmit queue, select buffer unit, the transmission wheel sub-value of supposing message to be sent is 1, represent that this message needs to take turns transmission the 2nd, after having found cacheable unit, check in the storage organization of this unit correspondence whether message is arranged, as do not have then directly message to be deposited in the storage organization of this unit correspondence, as be checked through in the storage organization of this unit correspondence and store message to be sent, whether then check this existing message identical with the transmission wheel sub-value of desiring to deposit in message, for example, suppose that it is 0 that this existing message sends the wheel sub-value, represent that this message needs to take turns transmission the 1st, since should existing message to take turns sub-value different with the transmission of desiring to deposit in message to be sent, then this can be desired to deposit in the storage organization that message deposits described unit correspondence in, finish caching, as being checked identical, transmission wheel sub-value such as this existing message also is 1, expression should also need to take turns transmission the 2nd by existing message, then the message to be sent that this can not be desired to deposit in deposits in, in order to avoid cause confusion, but continue the search buffer unit to deposit this message to be sent in.
Step 703, finish deposit in after, when sending according to the strategy scheduling message of setting, when being dispatched to the unit that stores message to be sent, the predetermined message of storing in this unit corresponding stored structure is sent, predetermined message described here is exactly that to send the wheel sub-value be 0 message, and other messages in this unit, then according to its transmission wheel sub-value separately afterwards round send.Be briefly described as follows, as sending the wheel sub-value in this unit is 0 message, then represent the message that this unit does not have needs to send immediately now, the scheduling that then continues lower unit sends, but no matter whether this unit has message to send, all need the transmission wheel sub-value of residue message in the described unit corresponding stored structure is subtracted one after having dispatched this unit, the promptly all corresponding previous round of carrying, just send the wheel sub-value and be 1 change 0 into, send the wheel sub-value and be 2 change 1 into, sending the wheel sub-value is the N-1 that changes into of N, so again through several take turns scheduling after, the message transmission that finally can be scheduled away.From foregoing description, be understood that under the certain similarity condition of transmit queue length, transmit queue of the present invention can be stored more message, and can send in correct, the utilization ratio of internal memory is improved greatly.
The above only is a preferred implementation of the present invention, does not constitute the qualification to protection range of the present invention.Any any modification of being done within the spirit and principles in the present invention, be equal to and replace and improvement etc., all should be included within the claim protection range of the present invention.
Claims (10)
1, a kind of flow shaping method in data transmission process is characterized in that: this method may further comprise the steps:
A) set up storage structure for storing waiting message, the corresponding described storage organization in each unit in the transmit queue;
B) determine the cacheable unit of message to be sent in the transmit queue, then packet buffer to be sent is gone in the storage organization of described cacheable unit correspondence;
C) the predetermined message according to buffer memory in each unit corresponding stored structure in the scheduling strategy scheduling transmit queue of setting sends.
2, flow shaping method according to claim 1 is characterized in that: described storage organization can be a chained list, an array or a first in first out (FIFO) formation.
3, flow shaping method according to claim 1 and 2, it is characterized in that: described method also comprises: according to the target transmitting time of message to be sent, current time and scheduling transmit queue one are taken turns the required time and are determined transmission round (roundcount) value of message to be sent, and should send and take turns sub-value and described message to be sent is made as one.
4, flow shaping method according to claim 3 is characterized in that: described will send wheel sub-value and message to be sent be made as one can be by in the heading of described message to be sent, being encapsulated into this data.
5, flow shaping method according to claim 3 is characterized in that: described will send wheel sub-value and message to be sent be made as one can be by taking turns sub-value in place, address this transmission of mark of depositing this message to be sent.
6, flow shaping method according to claim 3 is characterized in that: the caching of the described message to be sent of step b) is realized according to following step:
(11) but in the unit of described transmit queue search buffer memory message to be sent;
(12) do not have in the storage organization as the unit correspondence that searches message to be sent or not with the identical message of transmission wheel sub-value of the message of desiring to deposit in to be sent, then this message to be sent is deposited in the storage organization of described unit correspondence;
As the identical message of transmission wheel sub-value of existing and the message of desiring to deposit in to be sent in the storage organization of the unit correspondence that searches, then repeat above-mentioned steps (11), (12) but continue to search for buffer unit to deposit this message to be sent in.
7, flow shaping method according to claim 3 is characterized in that: the described predetermined message of step c) is that transmission wheel sub-value is 0 message.
8, flow shaping method according to claim 3 is characterized in that: step c) also comprises has dispatched in the transmit queue behind the unit, and sub-value is corresponding subtracts one with the transmission wheel of residue message in this unit.
9, flow shaping method according to claim 1 is characterized in that: the pointer that points to the storage organization corresponding with it by storage in each unit of transmit queue is realized corresponding between the corresponding storage organization of each unit and its.
10, flow shaping method according to claim 1 is characterized in that: realize corresponding between the corresponding storage organization of each unit and its by the mark of the storage storage organization corresponding with it in each unit of transmit queue.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB031092004A CN100495974C (en) | 2003-04-07 | 2003-04-07 | Flow shaping method in data transmission process |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB031092004A CN100495974C (en) | 2003-04-07 | 2003-04-07 | Flow shaping method in data transmission process |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1536816A true CN1536816A (en) | 2004-10-13 |
CN100495974C CN100495974C (en) | 2009-06-03 |
Family
ID=34319234
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB031092004A Expired - Fee Related CN100495974C (en) | 2003-04-07 | 2003-04-07 | Flow shaping method in data transmission process |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100495974C (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN100466603C (en) * | 2005-05-17 | 2009-03-04 | 华为技术有限公司 | Method and apparatus for shaping transmission service stream in network |
CN100466629C (en) * | 2006-09-18 | 2009-03-04 | 杭州华三通信技术有限公司 | Network equipment and message transferring method based on multiple-core processor |
CN102217256A (en) * | 2011-05-26 | 2011-10-12 | 华为技术有限公司 | Queue scheduling method and apparatus |
CN109218221A (en) * | 2017-06-30 | 2019-01-15 | 深圳市中兴微电子技术有限公司 | Traffic shaping resource management system and method |
-
2003
- 2003-04-07 CN CNB031092004A patent/CN100495974C/en not_active Expired - Fee Related
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN100466603C (en) * | 2005-05-17 | 2009-03-04 | 华为技术有限公司 | Method and apparatus for shaping transmission service stream in network |
CN100466629C (en) * | 2006-09-18 | 2009-03-04 | 杭州华三通信技术有限公司 | Network equipment and message transferring method based on multiple-core processor |
CN102217256A (en) * | 2011-05-26 | 2011-10-12 | 华为技术有限公司 | Queue scheduling method and apparatus |
WO2011144093A3 (en) * | 2011-05-26 | 2012-05-03 | 华为技术有限公司 | Method and device for scheduling queue |
CN102217256B (en) * | 2011-05-26 | 2015-03-18 | 华为技术有限公司 | Queue scheduling method and apparatus |
CN109218221A (en) * | 2017-06-30 | 2019-01-15 | 深圳市中兴微电子技术有限公司 | Traffic shaping resource management system and method |
CN109218221B (en) * | 2017-06-30 | 2022-05-27 | 深圳市中兴微电子技术有限公司 | Traffic shaping resource management system and method |
Also Published As
Publication number | Publication date |
---|---|
CN100495974C (en) | 2009-06-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11709702B2 (en) | Work conserving, load balancing, and scheduling | |
CN1064500C (en) | Method and apparatus for temporarily storing data packets | |
US7957392B2 (en) | Method and apparatus for high-performance bonding resequencing | |
WO2012162949A1 (en) | Packet reassembly and resequence method, apparatus and system | |
CN101136854B (en) | Method and apparatus for implementing data packet linear speed processing | |
CN112084136B (en) | Queue cache management method, system, storage medium, computer device and application | |
US8910171B2 (en) | Thread synchronization in a multi-thread network communications processor architecture | |
CN1736066A (en) | State engine for data processor | |
US20150331718A1 (en) | Network processor having multicasting protocol | |
CN1261486A (en) | Round-trip delay optimization of multiplexed speech channels | |
US7352766B2 (en) | High-speed memory having a modular structure | |
CN1859325A (en) | News transfer method based on chained list process | |
CN101035059A (en) | Method for improving the classification searching speed of the three-folded content addressable memory message | |
US7126959B2 (en) | High-speed packet memory | |
AU2004310639B2 (en) | Using ordered locking mechanisms to maintain sequences of items such as packets | |
CN1829200A (en) | Systems and methods for implementing counters in a network processor | |
CN1536816A (en) | Flow shaping method in data transmission process | |
CN1378143A (en) | Method for realizing quick data transfer | |
CN1499786A (en) | Group broadcast system and system used in linked node of ATM exchanger | |
US20090182714A1 (en) | Sorting apparatus and method | |
CN1499751A (en) | Data transmitting device and method for transmitting and receiving data and data communication system | |
CN101964751A (en) | Transmission method and device of data packets | |
CN1669279A (en) | Increasing memory access efficiency for packet applications | |
CN1620063A (en) | Method, system, and apparatus for a credit based flow control in a computer system | |
CN1210956C (en) | Real time receiving and sxtorage method for video signal conference stream medium |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090603 Termination date: 20170407 |