CN103888452B - For the order-preserving method and device of message compression - Google Patents
For the order-preserving method and device of message compression Download PDFInfo
- Publication number
- CN103888452B CN103888452B CN201410086621.6A CN201410086621A CN103888452B CN 103888452 B CN103888452 B CN 103888452B CN 201410086621 A CN201410086621 A CN 201410086621A CN 103888452 B CN103888452 B CN 103888452B
- Authority
- CN
- China
- Prior art keywords
- message
- compressed
- compression
- queue
- request queue
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Computer And Data Communications (AREA)
Abstract
The invention discloses a kind of order-preserving method and device for message compression, wherein the method includes:Multiple messages are received, and judges whether each message meets contractive condition successively;The message to be compressed for meeting contractive condition is sent to compression request queue, and the message non-to be compressed for not meeting contractive condition is sent to interim buffer queue;Whenever message to be compressed is sent to before compression request queue, in judging interim buffer queue, current whether there is has non-message to be compressed;If it is present non-message to be compressed and message to be compressed are sent in the corresponding memory element of respective meta-data into compression request queue;And to compression request queue in message to be compressed be compressed process, by compress response queue compression is processed after message to be compressed and metadata in message non-to be compressed sequentially export.The order-preserving method of the embodiment of the present invention effectively utilizes compression coprocessor bandwidth resources, optimizes compressing card I/O bandwidth and CPU service efficiencies.
Description
Technical field
The present invention relates to network communication technology field, more particularly to a kind of order-preserving method and device for message compression.
Background technology
At present, Web compression is generally based on the data compression of the application layer on transmission control protocol TCP, such as hypertext
Transfer protocol HTTP, it is main to be controlled to TCP layer by fine granularity to realize the compression to business datum, but to business datum
Transparent compression not yet finds.For example, the acceleration system using coprocessor or central processor CPU is similar to, FIFO can be passed through
(First in First Out, First Input First Output) mechanism realizes the communication between CPU and coprocessor, while by being based on
The paralleling tactic of TCP flow granularity, to guarantee the message in TCP flow without out of order generation.
The problem that presently, there are is, the application of the realization of FIFO mechanism for above-mentioned this similar compression, can will be all of
Business datum is all processed through coprocessor, can so cause the width of data compression coprocessor to waste problem, while
Also additionally consume unnecessary input/output port (I/O) width.
The content of the invention
It is contemplated that at least solving one of technical problem in correlation technique to a certain extent.
For this purpose, first purpose of the present invention is to propose a kind of order-preserving method for message compression.The method can be with
Effectively utilize compression coprocessor bandwidth resources, it is also possible to reduce unnecessary coprocessor I/O and access, optimization CPU is used
Efficiency, and by adaptive order-preserving push-mechanism, reduce the accumulation process delay during non-packet order preserving to be compressed.
Second object of the present invention is to propose a kind of order-preserving device for message compression.
To achieve these goals, the order-preserving method for message compression of first aspect present invention embodiment, including:Connect
Multiple messages are received, and judges whether each described message meets contractive condition successively;The to be compressed of the contractive condition will be met
Message is sent to compression request queue, and the message non-to be compressed for not meeting the contractive condition is sent to interim buffering team
Row;Whenever the message to be compressed is sent to before the compression request queue, judge current in the interim buffer queue
With the presence or absence of there is non-message to be compressed;If there is currently the non-message to be compressed, by the non-message to be compressed and
The message to be compressed is sent in the corresponding memory element of respective meta-data into the compression request queue;And to described
Message described to be compressed in compression request queue is compressed process, and treating after processed by compression is pressed
Described non-message to be compressed in contracting message and the metadata is sequentially exported into next stage processing system or transmission network.
The order-preserving method for message compression of the embodiment of the present invention, whenever message to be compressed is sent to compression request team
Before row, in judging interim buffer queue, current whether there is has non-message to be compressed, if there is currently non-report to be compressed
Text, then non-message to be compressed and message to be compressed are sent the corresponding memory element of respective meta-data into compression request queue
In, and after the message to be compressed in compression request queue is compressed process, by compressing response queue by compression
The message non-to be compressed in message to be compressed and metadata after reason is sequentially exported into next stage processing system or transmission network,
At least there is advantages below:(1) data compression is carried out based on TCP Flow granularities, different TCP Flow are allowed using different pressures
Compression method, realizes the data compression transparent to business datum;(2) data compression is carried out by compression coprocessor, association is processed
Asynchronous operation is carried out between device and CPU, the systematic function of compression device is improve;(3) by being less than preset value to message length
Message filtered, it is to avoid small messages compression improves the payload of data compression coprocessor without compression income
The I/O performances of utilization rate and optimization coprocessor.
To achieve these goals, the order-preserving device for message compression of second aspect present invention embodiment, including:Connect
Module is received, for receiving multiple messages;First judge module, for judging whether each described message meets compressor bar successively
Part;Sending module, for the message to be compressed for meeting the contractive condition is sent to compression request queue, and will not meet institute
The message non-to be compressed for stating contractive condition is sent to interim buffer queue;Second judge module, for whenever the sending module
The message to be compressed is sent to before the compression request queue, currently be whether there is in judging the interim buffer queue
There is non-message to be compressed;The sending module is additionally operable in second judge module judges the interim buffer queue currently
When there are the non-message to be compressed, the non-message to be compressed and the message to be compressed are sent to the compression request
In the corresponding memory element of respective meta-data in queue;Compression order-preserving module, for the institute in the compression request queue
State message to be compressed and be compressed process, the message to be compressed and the metadata after processed by compression
In described non-message to be compressed sequentially export into next stage processing system or transmission network.
The order-preserving device for message compression of the embodiment of the present invention, whenever sending module sends message to be compressed to pressure
Before contracting request queue, judge that current whether there is has non-message to be compressed in interim buffer queue by the second judge module,
Sending module sends non-message to be compressed and message to be compressed to compression request team when there is currently non-message to be compressed
In the corresponding memory element of respective meta-data in row, compression to be compressed message of the order-preserving module in compression request queue enters
The report non-to be compressed in message to be compressed and metadata after row compression is processed, after processed by compression
Text is sequentially exported into next stage processing system or transmission network, at least with advantages below:(1) entered based on TCP Flow granularities
Row data compression, different TCP Flow are allowed using different compression methods, realize the data compression transparent to business datum;
(2) data compression is carried out by compression coprocessor, asynchronous operation is carried out between coprocessor and CPU, compression device is improve
Systematic function;(3) by filtering less than the message of preset value to message length, it is to avoid small messages compressions is received without compression
Benefit, improves the I/O performances of the utilization rate and optimization coprocessor of the payload of data compression coprocessor.
The additional aspect of the present invention and advantage will be set forth in part in the description, and partly will become from the following description
Obtain substantially, or recognized by the practice of the present invention.
Description of the drawings
Of the invention above-mentioned and/or additional aspect and advantage will become from the following description of the accompanying drawings of embodiments
It is substantially and easy to understand, wherein,
Fig. 1 is the flow chart of the order-preserving method for message compression according to an embodiment of the invention;
Fig. 2 is the flow chart for judging that message meets contractive condition according to an embodiment of the invention;
Fig. 3 is the schematic diagram of the order-preserving method for message compression according to an embodiment of the invention;
Fig. 4 is schematic diagram of the non-message to be compressed according to an embodiment of the invention more than the TCP flow of message to be compressed;
Fig. 5 is the flow chart that the order-preserving method to for message compression according to an embodiment of the invention is optimized;
Fig. 6 is the schematic diagram that the order-preserving method to for message compression according to an embodiment of the invention is optimized;
Fig. 7 is the structural representation of the order-preserving device for message compression according to an embodiment of the invention;
Fig. 8 is the structural representation of the order-preserving device for message compression according to a specific embodiment of the invention.
Specific embodiment
Embodiments of the invention are described below in detail, the example of the embodiment is shown in the drawings, wherein from start to finish
Same or similar label represents same or similar element or the element with same or like function.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to for explaining the present invention, and be not considered as limiting the invention.
Following disclosure provides many different embodiments or example is used for realizing the different structure of the present invention.For letter
Change disclosure of the invention, hereinafter the part and setting of specific examples are described.Certainly, they are only merely illustrative, and
Purpose does not lie in the restriction present invention.Additionally, the present invention can in different examples repeat reference numerals and/or letter.It is this heavy
It is the relation between itself not indicating discussed various embodiments and/or arranging for purposes of simplicity and clarity again.
Below with reference to the accompanying drawings order-preserving method and device for message compression according to embodiments of the present invention described.
A kind of order-preserving method for message compression provided in an embodiment of the present invention, including:Receive multiple messages, and successively
Judge whether each message meets contractive condition;The message to be compressed for meeting contractive condition is sent to compression request queue, and
The message non-to be compressed for not meeting contractive condition is sent to interim buffer queue;Please to compression whenever message to be compressed is sent
Before asking queue, in judging interim buffer queue, current whether there is has non-message to be compressed;Non- wait to press if there is currently
Contracting message, then non-message to be compressed and message to be compressed are sent the corresponding storage of respective meta-data into compression request queue
In unit;And to compression request queue in message to be compressed be compressed process, by compress response queue by compression
The message non-to be compressed in message to be compressed and metadata after reason is sequentially exported into next stage processing system or transmission network.
Fig. 1 is the flow chart of the order-preserving method for message compression according to an embodiment of the invention.
As shown in figure 1, the order-preserving method for being used for message compression can include:
S101, receives multiple messages, and judges whether each message meets contractive condition successively.
Specifically, TCP (Transmission Control Protocol, transmission control protocol) stream (TCP can first be received
Flow the multiple messages included in), can judge whether each message meets contractive condition afterwards successively, to meet compressor bar
The message of part is compressed, and the message for being unsatisfactory for contractive condition is not compressed.
In one embodiment of the invention, as shown in Fig. 2 judging that message meets contractive condition and can include:
S201, obtains the length of each message.
Whether S202, judge the length of each message more than or equal to preset value.
Wherein, in an embodiment of the present invention, preset value can be by default.For example, system can be according to message length
Practical situation determines the size of the preset value, can collect a large amount of messages, first obtains the length of these messages, afterwards to these length
It is averaging, to obtain meansigma methodss, preset value can be determined according to the meansigma methodss afterwards.
S203, if greater than or be equal to preset value, then judge that message meets contractive condition.
S204, if less than preset value, then judges that message does not meet contractive condition.
Thus, can judge whether the message meets contractive condition according to the length of message, improve the accurate of judged result
Degree.
For example, as shown in figure 3, continued presence P0 in the TCP Flow for receiving, P1 ... P6 this seven messages, according to
The secondary length for obtaining this seven messages, can determine whether the limited length of outgoing packet P2, P4, P5, i.e., less than preset value, that is to say, that
Message P2, P4, P5 are the message for not meeting contractive condition, can be without the need for being compressed to which;Judge outgoing packet P0, P1, P3, P6
Length is the message for meeting contractive condition more than preset value, i.e. message P0, P1, P3, P6, and which need to be compressed.
S102, the message to be compressed for meeting contractive condition is sent to compression request queue, and will not meet contractive condition
Message non-to be compressed send to interim buffer queue.
For example, as shown in figure 3, being compressed caused association to optimize the message without compression income this to P2, P4, P5
Processor expense, it is to avoid message P2, P4, P5 enter I/O (Input/Output, the input/output port) stages.Therefore, work as process
During to message P3, these three messages to be compressed of total P0, P1 and P3 in compression request queue, message P2 will be filtered out.Now, such as
Message P2 is directly transparent to next stage Message processing process or output to network interface card by fruit, then may cause message P2 in message
Export before P0 and P1 in transmission network, so as to cause out of order generation.
For this purpose, in an embodiment of the present invention, can by message P0, P1, P3, P6 to be compressed for meeting contractive condition send to
Compression request queue, and message P2, P4, P5 non-to be compressed for not meeting contractive condition are sent into interim buffer queue.Its
In, as shown in figure 3, Order_buffer can be used to control out of order interim buffer queue in TCP Flow session structures
In the non-message to be compressed of blotter.Additionally, non-message P2, P4, P5 to be compressed are without the need for compression, and tail of the queue enqueue operations can be passed through
Recorded in Order_buffer, and can bypassing compression system and export in rear class system.
S103, whenever message to be compressed is sent to before compression request queue, in judging interim buffer queue be currently
It is no to there are non-message to be compressed.
For example, as shown in figure 3, compression engine judges that message P1 meets contractive condition, when by message P1 to be compressed send to
Before compression request queue, in can first judging now interim buffer queue, current whether there is has non-message to be compressed.
S104, if there is currently no non-message to be compressed, by the message to be compressed for meeting contractive condition send to
In the corresponding memory element of respective meta-data in compression request queue.
For example, when message P1 to be compressed is sent to before compression request queue, judge in now interim buffer queue when
It is front not there are non-message to be compressed, the message P1 to be compressed can be sent the respective meta-data into compression request queue afterwards
(metadata) in corresponding memory element.
S105, if there is currently non-message to be compressed, non-message to be compressed and message to be compressed is sent to pressure
In the corresponding memory element of respective meta-data in contracting request queue.
Specifically, when non-message to be compressed has been there is currently in interim buffer queue is judged, can first extract interim slow
The message non-to be compressed that there is currently in queue is rushed, the message to be compressed and the non-of extraction that meet contractive condition can be waited to press afterwards
Contracting message is sent into compression request queue together.Wherein, in an embodiment of the present invention, non-message to be compressed can be with pointer
Form is recorded into the corresponding metadata of message to be compressed.Additionally, in one embodiment of the invention, metadata has to be waited to press
The storage address of contracting message.That is, whenever it there are message enqueue operations to be compressed, extracting the interim chains of oder_buffer
What team's head of table was indicated bypasses message and recorded in the metadata of compression queue continuous item.
For example, as shown in figure 3, compression engine judges that message P2 is unsatisfactory for contractive condition, can first by the non-message to be compressed
P2 is sent into interim buffer queue.Afterwards, compression engine judges that message P3 meets contractive condition, when by message P3 to be compressed
Before delivering to compression request queue, in judging now interim buffer queue, non-message P2 to be compressed has been there is currently.Afterwards can be first
The message P2 non-to be compressed that there is currently in extracting interim buffer queue, and by team's head pointer of Order_buffer interim chained list
It is empty.Finally, message P3 to be compressed and non-message P2 to be compressed can be sent into compression request queue together.
And for example, as shown in figure 3, compression engine judges that message P6 meets contractive condition, when by message P6 to be compressed send to
During compression request queue, message P6 to be compressed can bring the message P4 and P5 non-to be compressed in interim buffer queue into compressions together
In request queue.
S106, to compression request queue in message to be compressed be compressed process, by compress response queue will compress
The message non-to be compressed in message to be compressed and metadata after process is sequentially exported to next stage processing system or transmission network
In.
Specifically, first can obtain message to be compressed according to the storage address in metadata, and treat compressed packet is carried out
Compression is processed.Afterwards, the storage address of the message to be compressed in metadata can be replaced with the message to be compressed after compression is processed
Storage address.Finally, after the completion of message to be compressed is compressed process and when being taken away from compression response queue, can be first
The pointer recorded in Sequential output each metadata, and beginning stepping through from the head of pointer, be sequentially output non-message to be compressed and
Message to be compressed after compression process.
For example, as shown in figure 3, at a time, there is message P3 to be compressed in compression request queue, while to be compressed
The non-message P2 to be compressed of the subsidiary pointer record of message P3;There is message P6 to be compressed in compression request queue, while to be compressed
After message P6 enters compression request queue, the associated pointers of team's head pointer of order_buffer to message P6 to be compressed are can record
In.
For example, when compression engine treats compressed packet P3 carries out dequeue operation, first can Sequential output it is currently to be compressed
The out of order pointer of the corresponding metadata records of message P3, travels through and is sequentially output non-message P2 to be compressed, most from the pointer head
After export message P3 to be compressed.Thus, it is possible to guarantee non-message P2, P4 and P5 to be compressed without actual compression process, no
Consume compression bandwidth, do not cause coprocessor I/O processes, while nor affecting on overall compression performance, can also finally reach
The order-preserving process of data message in TCP Flow.
The order-preserving method for message compression of the embodiment of the present invention, whenever message to be compressed is sent to compression request team
Before row, in judging interim buffer queue, current whether there is has non-message to be compressed, if there is currently non-report to be compressed
Text, then non-message to be compressed and message to be compressed are sent the corresponding memory element of respective meta-data into compression request queue
In, and after the message to be compressed in compression request queue is compressed process, by compressing response queue by compression
The message non-to be compressed in message to be compressed and metadata after reason is sequentially exported into next stage processing system or transmission network,
At least there is advantages below:(1) data compression is carried out based on TCP Flow granularities, different TCP Flow are allowed using different pressures
Compression method, realizes the data compression transparent to business datum;(2) data compression is carried out by compression coprocessor, association is processed
Asynchronous operation is carried out between device and CPU, the systematic function of compression device is improve;(3) by being less than preset value to message length
Message filtered, it is to avoid small messages compression improves the payload of data compression coprocessor without compression income
The I/O performances of utilization rate and optimization coprocessor.
It should be noted that inventor is analyzed by the data message to actual data transfer, discovery has kind of a phenomenon
Generally existing, for example, in the transmitting procedure of Web Page webpages, a direction overwhelming majority is little ACK
(Acknowledgement confirms character) message, these ACK messages do not meet contractive condition, and meet treating for contractive condition
The distribution substantially sparse distribution in non-message to be compressed of compressed packet.According to patrolling for the above-mentioned order-preserving method for message compression
Volume, if message to be compressed is not slowly admitted to compression request queue, the message non-to be compressed in interim buffer queue will one
Directly wait the message to be compressed to be carried in compression request queue, will so occur that non-message accumulation to be compressed is too high
The phenomenon of delay, or even affect the correctness of TCP performances and TCP upper-layer protocols.
For example, as shown in figure 4, the openness contrast of non-message to be compressed and message to be compressed, in figure, ACK is represented
ACK messages and incidentally very small amount TCP payload or the message without TCP payload, this phenomenon are prevalent in bulk
In the data transmission procedure of the unidirectional push of data.
The message big in order to process this sparse sex differernce for greater flexibility, is optimized design to above-mentioned order-preserving method.
In embodiments of the invention, for non-message to be compressed is significantly more than the scene of message to be compressed, it is to avoid by message to be compressed
To push, but by the direct active push output of non-message to be compressed;Must assure that before the non-message to be compressed of active push
The message not having in compressing card in current session, otherwise may be out of order, so all current in needing to perceive compressing card queue
The number of session messages.Specifically, in one embodiment of the invention, as shown in figure 5, being used for message compression to this
Order-preserving method is optimized can be included:
S501, obtains the number of current message to be compressed in compression request queue.
For example, can be the session addition Fly_in_HW_cnt variables in each message, the variable can be represented works as
The number of the front clear text in compressing card, can safeguard and currently be still within compression request queue and compression response queue
All messages.For example, when a message to be compressed enters compression request queue, can perform Fly_in_HW_cnt++ behaviour
Make;When message to be compressed after a compression is processed goes out to compress response queue, Fly_in_HW_cnt-- operations are can perform.
That is, can pass through to obtain the current value of Fly_in_HW_cnt variables, you can current to be compressed in acquisition compression request queue
The number of message.
S502, the message to be compressed after compression is processed are gone out in team, and compression request queue by compressing response queue
When the number of current message to be compressed is changed into pre-set threshold value, by the message output current all non-to be compressed in interim buffer queue
Into next stage processing system or transmission network.
Wherein, in an embodiment of the present invention, pre-set threshold value can be 0.For example, as shown in fig. 6, when going out group, working as Fly_in_
When in HW_cnt--=0, i.e. compression request queue, last message to be compressed goes out group, while interim buffer queue Order_
When currently also there are non-message to be compressed in buffer, now, can be by Flush Order_buffer by force by Order_
In buffer, all non-message to be compressed of current record is directly output to next stage and processes or export network interface card, needs not move through
Compressing card queue.Wherein, Flush Order_buffer can safeguard all interim non-messages to be compressed.Thus, it is to avoid deadlock
And high latency, optimize performance.
S503, when the message for judging to be currently received is pressed as non-message to be compressed, and currently treating in compression request queue
When the number of contracting message is equal to pre-set threshold value, non-message to be compressed is directly output to into next stage processing system or transmission network
In.
For example, as shown in fig. 6, when joining the team, working as Fly_in_HW_cnt==0, i.e., sent currently without message to be compressed
During in compression request queue, while when there is currently no non-message to be compressed in interim buffer queue Order_buffer,
Now, the message non-to be compressed for not meeting contractive condition in TCP Flow needs not move through Order_buffer bufferings, Ke Yizhi
Tap into into next stage process or export compressing card.So, deadlock and high latency can be avoided, optimizes performance.
In sum, this optimisation strategy is fully from single session granularities angle, it is contemplated that non-message to be compressed exists
Accumulated delays control in order_buffer, i.e., for the stream non-to be compressed of continuous big section, can complete transparent transmission with adaptive
Function.That is, for non-message to be compressed, can direct bypassing compression card hardware handles flow process, optimize compressing card I/O
Bandwidth and CPU service efficiencies, it is to avoid the compression bandwidth wasting of resources, also, as Fly_in_HW_cnt--==0, pass through
Flush Order_buffer by force by the message non-to be compressed in interim buffer queue be directly output to next stage process or
Export to network interface card, reduce the accumulation process delay during non-packet order preserving to be compressed.
From above-described embodiment, the number and interim buffer queue of current message to be compressed in compression request queue are obtained
In current non-message to be compressed number, when the number of the message current to be compressed in compression request queue is equal to pre-set threshold value,
And the number of the current non-message to be compressed in interim buffer queue is when being not equal to pre-set threshold value, by working as in interim buffer queue
Front non-message to be compressed is exported into next stage processing system or transmission network;When the report current to be compressed in compression request queue
When the number of the current non-message to be compressed that the number of text is equal in pre-set threshold value, and interim buffer queue is equal to pre-set threshold value,
The message non-to be compressed for not meeting contractive condition is directly output in next stage processing system or transmission network, compression is optimized
Card I/O bandwidth and CPU service efficiencies, it is to avoid the compression bandwidth wasting of resources, and reduce non-packet order preserving process to be compressed
In accumulation process postpone.
In order to realize above-described embodiment, the present invention also proposes a kind of order-preserving device for message compression.
A kind of order-preserving device for message compression that the embodiment of the present invention is proposed, including:Receiver module is more for receiving
Individual message;First judge module, for judging whether each message meets contractive condition successively;Sending module, for meeting
The message to be compressed of contractive condition is sent to compression request queue, and by the message non-to be compressed for not meeting contractive condition send to
Interim buffer queue;Second judge module, for message to be compressed being sent to before compression request queue whenever sending module,
In judging interim buffer queue, current whether there is has non-message to be compressed;Sending module is additionally operable to judge in the second judge module
When there is currently non-message to be compressed in interim buffer queue, non-message to be compressed and message to be compressed are sent please to compression
In seeking the corresponding memory element of respective meta-data in queue;Compression order-preserving module, presses for treating in compression request queue
Contracting message is compressed process, by compressing response queue and processing compression after message to be compressed and metadata in non-wait to press
Contracting message is sequentially exported into next stage processing system or transmission network.
Fig. 7 is the structural representation of the order-preserving device for message compression according to an embodiment of the invention.
As shown in fig. 7, the order-preserving device for being used for message compression can include:Receiver module 100, the first judge module
200th, sending module 300, the second judge module 400 and compression order-preserving module 500.
Specifically, receiver module 100 is used to receive multiple messages.First judge module 200 is used to judge each report successively
Whether text meets contractive condition.More specifically, multiple messages that receiver module 100 is included in can receiving TCP flow, afterwards, first
Judge module 200 can judge whether each message meets contractive condition successively, so that the message for meeting contractive condition is pressed
Contracting, the message for being unsatisfactory for contractive condition is not compressed.
In one embodiment of the invention, the first judge module 200 is additionally operable to each that acquisition receiver module 100 is received
The length of message, and whether the length of each message is judged more than or equal to preset value, and when more than or equal to preset value,
Judge that message meets contractive condition.Wherein, in an embodiment of the present invention, preset value can be by default.For example, system can root
The size of the preset value is determined according to the practical situation of message length, can collect a large amount of messages, first obtain the length of these messages, it
Afterwards these length are averaging, to obtain meansigma methodss, preset value can be determined according to the meansigma methodss afterwards.
In another embodiment of the present invention, the first judge module 200 is additionally operable to little in the length for judging each message
When preset value, can determine whether that message does not meet contractive condition.
Thus, can judge whether the message meets contractive condition according to the length of message, improve the accurate of judged result
Degree.
For example, as shown in figure 3, continued presence P0 in the TCP Flow that receive of receiver module 100, P1 ... P6 this
Seven messages, the first judge module 200 can obtain the length of this seven messages successively, and judge the length of outgoing packet P2, P4, P5
It is limited, i.e., less than preset value, that is to say, that message P2, P4, P5 are the message for not meeting contractive condition, can be without the need for carrying out to which
Compression;The length for judging outgoing packet P0, P1, P3, P6 is to meet contractive condition more than preset value, i.e. message P0, P1, P3, P6
Message, need to be compressed to which.
Sending module 300, and will be less than for the message to be compressed for meeting contractive condition is sent to compression request queue
The message non-to be compressed of sufficient contractive condition is sent to interim buffer queue.
For example, as shown in figure 3, being compressed caused association to optimize the message without compression income this to P2, P4, P5
Processor expense, it is to avoid message P2, P4, P5 enter the I/O stages.Therefore, when message P3 is processed, in compression request queue altogether
There are P0, P1 and P3 these three messages to be compressed, message P2 will be filtered out.Now, if message P2 is transparent to next stage directly
Message processing process or output may then cause message P2 that transmission network was exported before message P0 and P1 to network interface card
In, so as to cause out of order generation.
For this purpose, in an embodiment of the present invention, sending module 300 will can meet the message P0 to be compressed of contractive condition, P1,
P3, P6 are sent to compression request queue, and message P2, P4, P5 non-to be compressed for not meeting contractive condition are sent to interim slow
Rush in queue.Wherein, as shown in figure 3, Order_buffer can be control in TCP Flow session structures it is out of order interim
Buffer queue, for the non-message to be compressed of blotter.Additionally, non-message P2, P4, P5 to be compressed are without the need for compression, and can pass through
Tail of the queue enqueue operations are recorded in Order_buffer, and can bypassing compression system and export in rear class system.
Second judge module 400 for message to be compressed being sent to before compression request queue whenever sending module 300,
In judging interim buffer queue, current whether there is has non-message to be compressed.For example, as shown in figure 3, in the first judge module 200
Judge that message P1 meets contractive condition, and send to before compression request queue in 300 message P1 to be compressed of sending module, second
During judge module 400 can first judge now interim buffer queue, current whether there is has non-message to be compressed.
In one embodiment of the invention, sending module 300 is additionally operable to judge interim buffering in the second judge module 400
When there is currently no non-message to be compressed in queue, the message to be compressed for meeting contractive condition can be sent to compression request team
In the corresponding memory element of respective meta-data in row.For example, sending message P1 to be compressed in sending module 300 please to compression
Before asking queue, the second judge module 400 there is currently no non-message to be compressed in judging now interim buffer queue, afterwards
It is corresponding that the message P1 to be compressed can be sent respective meta-data (metadata) into compression request queue by sending module 300
In memory element.
In one embodiment of the invention, sending module 300 is additionally operable to judge interim buffering in the second judge module 400
When there is currently non-message to be compressed in queue, non-message to be compressed and message to be compressed are sent into compression request queue
The corresponding memory element of respective meta-data in.
More specifically, in an embodiment of the present invention, currently deposit in the second judge module 400 judges interim buffer queue
When there is non-message to be compressed, the message non-to be compressed that sending module 300 can be there is currently in first extracting interim buffer queue, it
The message to be compressed and the message non-to be compressed for extracting that meet contractive condition can be sent into compression request queue together afterwards.Its
In, in an embodiment of the present invention, non-message to be compressed can record metadata corresponding to message to be compressed with pointer form
In.Additionally, in one embodiment of the invention, metadata has the storage address of message to be compressed.That is, whenever depositing
In message enqueue operations to be compressed, sending module 300 can extract the bypass report that team's head of the interim chained lists of oder_buffer is indicated
In metadata that is literary and recorded compression queue continuous item.
For example, as shown in figure 3, when the first judge module 200 judges that message P2 is unsatisfactory for contractive condition, sending module
300 first can send the non-message P2 to be compressed into interim buffer queue.Afterwards, the first judge module 200 judges message P3
Meet contractive condition, message P3 to be compressed is sent to before compression request queue in sending module 300, the second judge module
400 judge now interim buffer queue in there is currently non-message P2 to be compressed.Sending module 300 can first extract interim buffering
The message P2 non-to be compressed that there is currently in queue, and team's head pointer of Order_buffer interim chained list is empty.Finally, may be used
Message P3 to be compressed and non-message P2 to be compressed are sent into compression request queue together.
And for example, as shown in figure 3, the first judge module 200 judges that message P6 meets contractive condition, will in sending module 300
Message P6 to be compressed is sent when compression request queue, and message P6 to be compressed can be by the message non-to be compressed in interim buffer queue
P4 and P5 are brought in compression request queue together.
Compression order-preserving module 500 is compressed processs for the message to be compressed in compression request queue, by compressing
Response queue compression is processed after message to be compressed and metadata in message non-to be compressed sequentially export to next stage process
In system or transmission network.
More specifically, in an embodiment of the present invention, compression order-preserving module 500 can first according to the storage address in metadata
Message to be compressed is obtained, and treats compressed packet and be compressed process.Afterwards, can be by the storage of the message to be compressed in metadata
Address replaces with the storage address of the message to be compressed after compression is processed.Finally, complete when message to be compressed is compressed process
When being taken away afterwards and from compression response queue, the pointer that can be recorded in first each metadata of Sequential output, and from the head of pointer
Portion begins stepping through, the message to be compressed after being sequentially output non-message to be compressed and compression process.
For example, as shown in figure 3, at a time, there is message P3 to be compressed in compression request queue, while to be compressed
The non-message P2 to be compressed of the subsidiary pointer record of message P3;There is message P6 to be compressed in compression request queue, while to be compressed
After message P6 enters compression request queue, the associated pointers of team's head pointer of order_buffer to message P6 to be compressed are can record
In.
For example, when compressing order-preserving module 500 and treating compressed packet P3 and carry out dequeue operation, first can Sequential output it is current
The out of order pointer of the corresponding metadata record of message P3 to be compressed, travels through and is sequentially output non-message to be compressed from the pointer head
P2, finally exports message P3 to be compressed.Thus, it is possible to guarantee non-message P2, P4 and P5 to be compressed without the compressed of reality
Journey, do not consume compression bandwidth, do not cause coprocessor I/O processes, while nor affecting on overall compression performance, finally can also
Reach the order-preserving process of data message in TCP Flow.
The order-preserving device for message compression of the embodiment of the present invention, whenever sending module sends message to be compressed to pressure
Before contracting request queue, judge that current whether there is has non-message to be compressed in interim buffer queue by the second judge module,
Sending module sends non-message to be compressed and message to be compressed to compression request team when there is currently non-message to be compressed
In the corresponding memory element of respective meta-data in row, compression to be compressed message of the order-preserving module in compression request queue enters
The report non-to be compressed in message to be compressed and metadata after row compression is processed, after processed by compression
Text is sequentially exported into next stage processing system or transmission network, at least with advantages below:(1) entered based on TCP Flow granularities
Row data compression, different TCP Flow are allowed using different compression methods, realize the data compression transparent to business datum;
(2) data compression is carried out by compression coprocessor, asynchronous operation is carried out between coprocessor and CPU, compression device is improve
Systematic function;(3) by filtering less than the message of preset value to message length, it is to avoid small messages compressions is received without compression
Benefit, improves the I/O performances of the utilization rate and optimization coprocessor of the payload of data compression coprocessor.
Fig. 8 is the structural representation of the order-preserving device for message compression according to a specific embodiment of the invention.
As shown in figure 8, the order-preserving device for being used for message compression can include:Receiver module 100, the first judge module
200th, sending module 300, the second judge module 400, compression order-preserving module 500 and the first acquisition module 600.
Specifically, the first acquisition module 600 is used for the number for obtaining current message to be compressed in compression request queue.Citing
For, the first acquisition module 600 can be that the session in each message adds Fly_in_HW_cnt variables, and the variable can be represented
The number of the current clear text in compressing card, can safeguard and currently be still within compression request queue and compression response queue
In all messages.For example, when a message to be compressed enters compression request queue, can perform Fly_in_HW_cnt++
Operation;When message to be compressed after a compression is processed goes out to compress response queue, Fly_in_HW_cnt-- behaviour is can perform
Make.That is, the first acquisition module 600 can pass through to obtain the current value of Fly_in_HW_cnt variables, you can obtaining compression please
Seek the number of current message to be compressed in queue.
In one embodiment of the invention, compress the message to be compressed that order-preserving module 500 is additionally operable to after compression is processed
When being changed into pre-set threshold value by compressing the number of the message current to be compressed that response queue goes out in team, and compression request queue, will
Current all non-message to be compressed in interim buffer queue is exported into next stage processing system or transmission network.Wherein, exist
In embodiments of the invention, pre-set threshold value can be 0.
For example, as shown in fig. 6, when going out group, in working as Fly_in_HW_cnt--=0, i.e. compression request queue, last is treated
When compressed packet goes out group, while when currently also there are non-message to be compressed in interim buffer queue Order_buffer, now,
The all non-of current record in Order_buffer can be treated by force by compression order-preserving module 500 by Flush Order_buffer
Compressed packet is directly output to next stage and processes or export network interface card, needs not move through compressing card queue.Wherein, Flush
Order_buffer can safeguard all interim non-messages to be compressed.Thus, it is to avoid deadlock and high latency, optimize performance.
In one embodiment of the invention, compress order-preserving module 500 to be additionally operable to judge currently to connect in the first judge module
The message for receiving is non-message to be compressed, and the number of the message current to be compressed in compression request queue is equal to pre-set threshold value
When, non-message to be compressed is directly output in next stage processing system or transmission network.
For example, as shown in fig. 6, when joining the team, working as Fly_in_HW_cnt==0, i.e., sent currently without message to be compressed
During in compression request queue, while when there is currently no non-message to be compressed in interim buffer queue Order_buffer,
Now, the message non-to be compressed for not meeting contractive condition in TCP Flow needs not move through Order_buffer bufferings, Ke Yizhi
Tap into into next stage process or export compressing card.So, deadlock and high latency can be avoided, optimizes performance.
In sum, this optimisation strategy is fully from single session granularities angle, it is contemplated that non-message to be compressed exists
Accumulated delays control in order_buffer, i.e., for the stream non-to be compressed of continuous big section, can complete transparent transmission with adaptive
Function.That is, for non-message to be compressed, can direct bypassing compression card hardware handles flow process, optimize compressing card I/O
Bandwidth and CPU service efficiencies, it is to avoid the compression bandwidth wasting of resources, also, as Fly_in_HW_cnt--==0, pass through
Flush Order_buffer by force by the message non-to be compressed in interim buffer queue be directly output to next stage process or
Export to network interface card, reduce the accumulation process delay during non-packet order preserving to be compressed.
The order-preserving device for message compression of the embodiment of the present invention, obtains compression request queue by the first acquisition module
In current message to be compressed number, the second acquisition module obtains the number of current non-message to be compressed in interim buffer queue,
The number of compression current to be compressed message of the order-preserving module in compression request queue is equal to pre-set threshold value, and interim buffer queue
In the number of current non-message to be compressed when being not equal to pre-set threshold value, by the current non-message to be compressed in interim buffer queue
Export into next stage processing system or transmission network;The number of the message current to be compressed in compression request queue is equal to pre-
If the number of the current non-message to be compressed in threshold values, and interim buffer queue is equal to pre-set threshold value, compressor bar will not met
The message non-to be compressed of part is directly output in next stage processing system or transmission network, optimizes compressing card I/O bandwidth and CPU
Service efficiency, it is to avoid the compression bandwidth wasting of resources, and reduce the accumulation during non-packet order preserving to be compressed and process and prolong
Late.
In flow chart or here any process described otherwise above or method description are construed as, expression includes
It is one or more for realizing specific logical function or process the step of the module of code of executable instruction, fragment or portion
Point, and the scope of the preferred embodiment of the present invention includes other realization, wherein the suitable of shown or discussion can not be pressed
Sequence, including according to involved function by it is basic simultaneously in the way of or in the opposite order, carry out perform function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or here logic described otherwise above and/or step, for example, are considered use in flow charts
In the order list of the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (as computer based system, the system including processor or other can hold from instruction
The system of row system, device or equipment instruction fetch execute instruction) use, or with reference to these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium " can any can be included, store, communicate, propagate or pass
The dress that defeated program is used for instruction execution system, device or equipment or with reference to these instruction execution systems, device or equipment
Put.The more specifically example (non-exhaustive list) of computer-readable medium is including following:With the electricity that one or more connect up
Connecting portion (electronic installation), portable computer diskette box (magnetic device), random access memory (RAM), read only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device, and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program thereon or other are suitable
Medium, because for example by carrying out optical scanning to paper or other media edlin, interpretation can then be entered or if necessary with which
His suitable method is processed to electronically obtain described program, is then stored in computer storage.
It should be appreciated that each several part of the present invention can be realized with hardware, software, firmware or combinations thereof.Above-mentioned
In embodiment, the software that multiple steps or method can be performed in memory and by suitable instruction execution system with storage
Or firmware is realizing.For example, if realized with hardware, and in another embodiment, can be with well known in the art
Any one of row technology or their combination are realizing:With for the logic gates of logic function is realized to data signal
Discrete logic, the special IC with suitable combinational logic gate circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
Those skilled in the art are appreciated that to realize all or part of step that above-described embodiment method is carried
Suddenly the hardware that can be by program to instruct correlation is completed, and described program can be stored in a kind of computer-readable storage medium
In matter, the program upon execution, including one or a combination set of the step of embodiment of the method.
Additionally, each functional unit in each embodiment of the invention can be integrated in a processing module, it is also possible to
It is that unit is individually physically present, it is also possible to which two or more units are integrated in a module.Above-mentioned integrated mould
Block both can be realized in the form of hardware, it would however also be possible to employ the form of software function module is realized.The integrated module is such as
Fruit using in the form of software function module realize and as independent production marketing or use when, it is also possible to be stored in a computer
In read/write memory medium.
Storage medium mentioned above can be read only memory, disk or CD etc..
Additionally, term " first ", " second " are only used for describing purpose, and it is not intended that indicating or implying relative importance
Or the implicit quantity for indicating indicated technical characteristic.Thus, define " first ", the feature of " second " can express or
Implicitly include one or more this feature.In describing the invention, " multiple " are meant that two or more,
Unless otherwise expressly limited specifically.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
Example ", or the description of " some examples " etc. mean specific features with reference to the embodiment or example description, structure, material or spy
Point is contained at least one embodiment or example of the present invention.In this manual, to the schematic representation of above-mentioned term not
Identical embodiment or example must be directed to.And, the specific features of description, structure, material or feature can be with office
Combined in one or more embodiments or example in an appropriate manner.Additionally, in the case of not conflicting, the skill of this area
The feature of the different embodiments or example described in this specification and different embodiments or example can be tied by art personnel
Close and combine.
Although embodiments of the invention have been shown and described above, it is to be understood that above-described embodiment is example
Property, it is impossible to limitation of the present invention is interpreted as, one of ordinary skill in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, changes, replacing and modification.
Claims (14)
1. a kind of order-preserving method for message compression, it is characterised in that include:
Multiple messages are received, and judges whether each described message meets contractive condition successively;
The message to be compressed for meeting the contractive condition is sent to compression request queue, and the contractive condition will not met
Non- message to be compressed is sent to interim buffer queue;
Whenever the message to be compressed is sent to before the compression request queue, judge current in the interim buffer queue
With the presence or absence of there is non-message to be compressed;
If there is currently the non-message to be compressed, by the non-message to be compressed and the message to be compressed send to
In the corresponding memory element of respective meta-data in the compression request queue;And
Process is compressed to the message described to be compressed in the compression request queue, by response queue being compressed by compression
Described non-message to be compressed in message to be compressed and the metadata after reason is sequentially exported to next stage processing system or biography
In defeated network.
2. method according to claim 1, it is characterised in that described that non-message to be compressed and the message to be compressed are sent out
The corresponding memory element of respective meta-data delivered in the compression request queue includes:
The described non-message to be compressed that there is currently in extracting the interim buffer queue;
The message described to be compressed and the described non-message to be compressed for extracting that meet the contractive condition are sent to the compression
In request queue, wherein, the non-message to be compressed is recorded into the corresponding metadata of the message to be compressed with pointer form.
3. method according to claim 1, it is characterised in that the metadata has the storage of the message to be compressed
Location, the message described to be compressed in the queue to compression request is compressed process to be included:
The message to be compressed is obtained according to the storage address, and process is compressed to the message to be compressed;
The storage address of the message described to be compressed in the metadata is replaced with into depositing for the message to be compressed after compression is processed
Storage address.
4. method according to claim 2, it is characterised in that it is described by compressing response queue and processing compression after treat
Described non-message to be compressed in compressed packet and the metadata is sequentially exported into next stage processing system or transmission network
Including:
The pointer recorded in metadata described in Sequential output, and the head from the pointer begins stepping through, and is sequentially output described non-
Message described to be compressed after message to be compressed and compression process.
5. method according to claim 1, it is characterised in that judge that the message meets the contractive condition and includes:
Obtain the length of each message;
Judge the length of each message whether more than or equal to preset value;
If greater than or be equal to the preset value, then judge that the message meets the contractive condition.
6. method according to claim 1, it is characterised in that also include:
Obtain the number of current message to be compressed in the compression request queue;
Message to be compressed after the compression is processed is by working as that compression response queue goes out in team, and the compression request queue
When the number of front message to be compressed is changed into pre-set threshold value, will be the current all non-message to be compressed in the interim buffer queue defeated
Go out into next stage processing system or transmission network.
7. method according to claim 6, it is characterised in that also include:
When the message for judging to be currently received is as the message current to be compressed in non-message to be compressed, and the compression request queue
Number be equal to the pre-set threshold value when, the non-message to be compressed is directly output to into next stage processing system or transmission network
In.
8. a kind of order-preserving device for message compression, it is characterised in that include:
Receiver module, for receiving multiple messages;
First judge module, for judging whether each described message meets contractive condition successively;
Sending module, for the message to be compressed for meeting the contractive condition is sent to compression request queue, and will not meet
The message non-to be compressed of the contractive condition is sent to interim buffer queue;
Second judge module, for whenever the sending module by the message to be compressed send to the compression request queue it
Before, in judging the interim buffer queue, current whether there is has non-message to be compressed;
The sending module is additionally operable to have there is currently in second judge module judges the interim buffer queue described
During non-message to be compressed, the non-message to be compressed and the message to be compressed are sent the phase into the compression request queue
In answering the corresponding memory element of metadata;
Compression order-preserving module, for being compressed process to the message described to be compressed in the compression request queue, by pressure
Contracting response queue compression is processed after message to be compressed and the metadata in described non-message to be compressed sequentially export to
In next stage processing system or transmission network.
9. device according to claim 8, it is characterised in that the sending module is additionally operable to:Extract the interim buffering
The described non-message to be compressed that there is currently in queue, and message described to be compressed and the extraction of the contractive condition will be met
The non-message to be compressed is sent into the compression request queue, wherein, the non-message to be compressed is recorded with pointer form
Into the corresponding metadata of the message to be compressed.
10. device according to claim 8, it is characterised in that the metadata has the storage of the message to be compressed
Address, the compression order-preserving module are additionally operable to:The message to be compressed is obtained according to the storage address, and to described to be compressed
Message is compressed process, and the storage address of the message described to be compressed in the metadata is replaced with after compression process
Message to be compressed storage address.
11. devices according to claim 9, it is characterised in that the compression order-preserving module specifically for:Sequential output institute
The pointer recorded in stating metadata, and the head from the pointer begins stepping through, and is sequentially output the non-message to be compressed and pressure
Message described to be compressed after contracting process.
12. devices according to claim 8, it is characterised in that first judge module is additionally operable to:Obtain the reception
The length of each message that module is received, and judge whether the length of each message is more than or equal to preset value, with
And when more than or equal to the preset value, judge that the message meets the contractive condition.
13. devices according to claim 8, it is characterised in that also include:
First acquisition module, for obtaining the number of current message to be compressed in the compression request queue;Wherein
The compression order-preserving module is additionally operable to the message to be compressed after the compression is processed and goes out team by compressing response queue, and
When the number of the message current to be compressed in the compression request queue is changed into pre-set threshold value, by the interim buffer queue
Current all non-messages to be compressed are exported into next stage processing system or transmission network.
14. devices according to claim 13, it is characterised in that the compression order-preserving module is additionally operable to sentence described first
Disconnected module judges the message that is currently received as the message current to be compressed in non-message to be compressed, and the compression request queue
Number be equal to the pre-set threshold value when, the non-message to be compressed is directly output to into next stage processing system or transmission network
In.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410086621.6A CN103888452B (en) | 2014-03-10 | 2014-03-10 | For the order-preserving method and device of message compression |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410086621.6A CN103888452B (en) | 2014-03-10 | 2014-03-10 | For the order-preserving method and device of message compression |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103888452A CN103888452A (en) | 2014-06-25 |
CN103888452B true CN103888452B (en) | 2017-04-05 |
Family
ID=50957172
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410086621.6A Active CN103888452B (en) | 2014-03-10 | 2014-03-10 | For the order-preserving method and device of message compression |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103888452B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109032506A (en) * | 2018-06-27 | 2018-12-18 | 郑州云海信息技术有限公司 | A kind of memory system data compression method, system and equipment and storage medium |
CN111459404B (en) * | 2020-03-03 | 2022-05-17 | 平安科技(深圳)有限公司 | Data compression method and device, computer equipment and storage medium |
CN112104658B (en) * | 2020-09-17 | 2022-08-12 | 山东方寸微电子科技有限公司 | Message compression method and system |
CN115134425B (en) * | 2022-06-20 | 2024-07-16 | 北京京东乾石科技有限公司 | Message processing method and device |
CN116155828B (en) * | 2022-12-21 | 2023-06-23 | 北京云豹创芯智能科技有限公司 | Message order keeping method and device for multiple virtual queues, storage medium and electronic equipment |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102571572A (en) * | 2011-12-28 | 2012-07-11 | 中国人民解放军国防科学技术大学 | Short message satellite communication-based data transmission method and device |
CN103369593A (en) * | 2012-04-05 | 2013-10-23 | 中兴通讯股份有限公司 | Method for compressing and decompressing Ethernet message and network element equipment |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7929562B2 (en) * | 2000-11-08 | 2011-04-19 | Genesis Telecommunications Laboratories, Inc. | Method and apparatus for optimizing response time to events in queue |
US7139794B2 (en) * | 2000-12-27 | 2006-11-21 | 3-D-V-U Israel (2000) Ltd. | System and methods for network image delivery with dynamic viewing frustum optimized for limited bandwidth communication channels |
-
2014
- 2014-03-10 CN CN201410086621.6A patent/CN103888452B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102571572A (en) * | 2011-12-28 | 2012-07-11 | 中国人民解放军国防科学技术大学 | Short message satellite communication-based data transmission method and device |
CN103369593A (en) * | 2012-04-05 | 2013-10-23 | 中兴通讯股份有限公司 | Method for compressing and decompressing Ethernet message and network element equipment |
Also Published As
Publication number | Publication date |
---|---|
CN103888452A (en) | 2014-06-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103888452B (en) | For the order-preserving method and device of message compression | |
EP3612942B1 (en) | Queue management for direct memory access | |
CN105653484B (en) | A kind of deblocking compression multi-channel transmission method | |
CN107257329B (en) | A kind of data sectional unloading sending method | |
JP5535115B2 (en) | Multithreaded file input / output system and multithreaded file input / output program | |
US20160132541A1 (en) | Efficient implementations for mapreduce systems | |
CN105183557B (en) | A kind of hardware based configurable data compression system | |
CN103412786A (en) | High performance server architecture system and data processing method thereof | |
JPWO2014061481A1 (en) | Data transfer apparatus and data transfer system using adaptive compression algorithm | |
CN103392169A (en) | Sorting | |
CN103294548A (en) | Distributed file system based IO (input output) request dispatching method and system | |
US9769241B2 (en) | Data transfer device, data transfer method, and program storage medium | |
CN103312336A (en) | Data compression method and device | |
CN103927263A (en) | Garbage recycling method and garbage recycling device | |
CN114201421A (en) | Data stream processing method, storage control node and readable storage medium | |
CN110191156B (en) | Data processing method and device based on block chain | |
US10673768B2 (en) | Managing data compression | |
JP4740828B2 (en) | Information processing apparatus and information processing system | |
CN109246331A (en) | A kind of method for processing video frequency and system | |
CN105701047B (en) | A kind of multi-channel data transmission and virtual restoration method | |
CN209803775U (en) | Data processing apparatus | |
CN107832256A (en) | A kind of method and device of data processing | |
CN102609217B (en) | High-speed storage method and high-speed storage system for IO (input/output) equipment | |
CN110532110A (en) | A kind of UART master control system for automatically switching to send data outside in multicore scene | |
CN104158834B (en) | A kind of method and apparatus that speech data is processed |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |