CN103929415B - The method, apparatus and network system of reading and writing data under RDMA - Google Patents
The method, apparatus and network system of reading and writing data under RDMA Download PDFInfo
- Publication number
- CN103929415B CN103929415B CN201410108028.7A CN201410108028A CN103929415B CN 103929415 B CN103929415 B CN 103929415B CN 201410108028 A CN201410108028 A CN 201410108028A CN 103929415 B CN103929415 B CN 103929415B
- Authority
- CN
- China
- Prior art keywords
- memory
- region
- address
- specified address
- rdma
- 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
- 238000000034 method Methods 0.000 title claims abstract description 73
- 230000015654 memory Effects 0.000 claims abstract description 339
- 238000012545 processing Methods 0.000 claims description 24
- 230000005540 biological transmission Effects 0.000 claims description 19
- 238000012790 confirmation Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims 2
- 238000000151 deposition Methods 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000003455 independent Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/061—Improving I/O performance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
- G06F3/0661—Format or protocol conversion arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/067—Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L61/00—Network arrangements, protocols or services for addressing or naming
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Bus Control (AREA)
Abstract
The present invention provides a kind of method that Remote Direct Memory accesses reading and writing data under RDMA, is related to computer realm, can support RDMA read operations and RDMA Write operation in the case where client reads or write scattered memory address to service end.Digital independent includes under RDMA：First end sends RDMA Read Request message to the second end, and the specified address for needing to read data is carried in the RDMA Read Request message, and the specified address is multiple addresses scattered in the range of the region of memory at second end；The first end receives data corresponding to specified address that second end is sent, described.The present invention also provides corresponding device.
Description
Technical field
The present invention relates to the communications field, more particularly to a kind of Remote Direct Memory accesses (remote direct
Memory access, RDMA) under reading and writing data method, apparatus and network system.
Background technology
In computer computing, RDMA refers to the direct storage of the memory from the memory of a computer to another computer
Device accesses, without regard to the operating system to any computer.RDMA allows the network connection of high flux, low delay, and this is big
The parallel PC cluster of scale is concentrated especially useful.
Three kinds of reading and writing data modes of RDMA technical supports at present：Transmission/reception (send/receive), RDMA are read
(RDMA read), RDMA Write enter (RDMA write).Wherein, scattered memory address is read or write to service end in client
In the case of, RDMA is read and RDMA Write operation is possible to be normally carried out.
The content of the invention
The embodiments of the invention provide a kind of method, apparatus and network system of RDMA Write reading and writing data, client to
In the case that service end is read or write-in disperses memory address, it would be preferable to support RDMA read operations and RDMA Write operation.
First aspect, there is provided a kind of method that Remote Direct Memory accesses digital independent under RDMA, methods described include：
First end sends RDMA Read Request message to the second end, and being carried in the RDMA Read Request message needs to read
The specified address of data, the specified address are multiple addresses scattered in the range of the region of memory at second end；
The first end receives data corresponding to specified address that second end is sent, described.
With reference in a first aspect, in the first implementation of first aspect, sent in the first end to the second end
Before RDMA Read Request message, methods described also includes：
The first end obtains the memory information of the region of memory at second end, wherein, the memory information includes：Ground
Location information and/or length information.
With reference to the first implementation of first aspect, in second of implementation of first aspect, second end
Region of memory include multiple scattered region of memorys, the first end obtains the memory information bag of the region of memory at second end
Include：
The first end obtains the memory information of each scattered region of memory in the region of memory at second end.
With reference to second of implementation of first aspect, in the third implementation of first aspect, it is described specifiedly
Location includes the multiple addresses being distributed in the multiple scattered region of memory, then
Data corresponding to specified address that first end reception second end is sent, described include：
Data corresponding to each address in specified address that first end reception second end is sent, described.
With reference in a first aspect, in the 4th kind of implementation of first aspect, sent in the first end to the second end
Before RDMA Read Request message, methods described also includes：
The first end is registered to the region of memory of itself to network interface card.
Second aspect, there is provided a kind of method that Remote Direct Memory accesses digital independent under RDMA, methods described include：
Second end receives the RDMA Read Request message that first end is sent, and being carried in the RDMA Read Request message needs
The specified address of data is read, the specified address is multiple addresses scattered in the range of the region of memory at second end；
Read data corresponding to the specified address in second end；
Second end sends data corresponding to the specified address to the first end.
With reference to second aspect, in the first implementation of second aspect, the specified address is read at second end
Corresponding data include：Read data corresponding to each address in the specified address in second end；
Data corresponding to second end to the first end transmission specified address include：Second end is to described
First end sends data corresponding to each address in the specified address.
With reference to second aspect, in second of implementation of second aspect, exist in second end multiple scattered interior
Region is deposited, the specified address includes the multiple addresses being distributed in the multiple scattered region of memory, then
Data corresponding to the second end reading specified address include：Read in the specified address at second end
Data corresponding to each address；
Data corresponding to second end to the first end transmission specified address include：Second end is to described
First end sends data corresponding to each address in the specified address.
With reference to second aspect, in the third implementation of second aspect, receive first end at second end and send
RDMA Read Request message before, methods described also includes：
The region of memory of the two port itself is registered to network interface card.
The third aspect, there is provided a kind of Remote Direct Memory accesses the method that data write under RDMA, and methods described includes：
Data corresponding to the specified address for needing write-in data and the specified address are encapsulated in RDMA Write and entered by first end
In data message, the specified address is multiple addresses scattered in the range of the region of memory at second end；
The first end sends the RDMA Write to second end and enters data message.
With reference to the third aspect, in the first implementation of the third aspect, will need to write data in the first end
Specified address and the specified address corresponding to data be encapsulated in before RDMA Write enters in data message, methods described is also wrapped
Include：
The first end obtains the memory information of the region of memory at second end, wherein, the internal memory of the region of memory
Information includes：The address information and/or length information of region of memory.
With reference to the first implementation of the third aspect, in second of implementation of the third aspect, second end
Region of memory include multiple scattered region of memorys, the first end obtains the memory information bag of the region of memory at second end
Include：
The first end obtains the memory information of each scattered region of memory in the region of memory at second end.
With reference to second of implementation of the third aspect, in the third implementation of the third aspect, it is described specifiedly
Location includes the multiple addresses being distributed in the multiple scattered region of memory.
With reference to the third aspect, in the 4th kind of implementation of the third aspect, sent out in the first end to second end
Send after the RDMA Write enters data message, methods described also includes：
The first end receives the confirmation message that the instruction that second end is sent is successfully written.
Fourth aspect, there is provided a kind of Remote Direct Memory accesses the method that data write under RDMA, and methods described includes：
Second end receives the RDMA Write that first end is sent and enters data message, and the RDMA Write enters carrying needs in data message
Data corresponding to the specified address and the specified address of write-in data, the specified address are the region of memory at second end
In the range of disperse multiple addresses；
Second end writes data corresponding to the specified address in the specified address.
With reference to fourth aspect, in the first implementation of fourth aspect, exist in second end multiple scattered interior
Region is deposited, the specified address includes the multiple addresses being distributed in the multiple scattered region of memory,
Second end includes in data corresponding to the specified address write-in：Second end is in the specified address
Each address write data corresponding to the address.
5th aspect, there is provided the device of digital independent, described device include under a kind of RDMA：
Sending module, for sending RDMA Read Request message to the second end, carried in the RDMA Read Request message
Need read data specified address, the specified address for second end region of memory in the range of disperse multiplely
Location；
Receiving module, for receiving data corresponding to specified address that second end is sent, described.
With reference to the 5th aspect, in the first implementation of the 5th aspect, described device also includes：
Acquisition module, for the sending module to the second end send RDMA Read Request message before, described in acquisition
The memory information of the region of memory at the second end, wherein, the memory information includes：Address information and/or length information.
With reference to the first implementation of the 5th aspect, in second of implementation of the 5th aspect, second end
Region of memory include multiple scattered region of memorys, then the acquisition module is specifically used for：
Obtain the memory information of each scattered region of memory in the region of memory at second end.
With reference to the 5th aspect second of implementation, the 5th aspect the third implementation in, it is described specifiedly
Location includes the multiple addresses being distributed in the multiple scattered region of memory, then
The receiving module is specifically used for：
Receive data corresponding to each address in specified address that second end is sent, described.
With reference to the 5th aspect, in the 4th kind of implementation of the 5th aspect, described device also includes：
Registering modules, for the sending module to the second end send RDMA Read Request message before, to itself
Region of memory is registered to network interface card.
A kind of 6th aspect, there is provided the device of digital independent under RDMA, it is characterised in that described device includes：
Receiving module, for receiving the RDMA Read Request message of first end transmission, in the RDMA Read Request message
The specified address for needing to read data is carried, the specified address is multiple to disperse in the range of the region of memory at second end
Address；
Processing module, for reading data corresponding to the specified address；
Sending module, for sending data corresponding to the specified address to the first end.
With reference to the 6th aspect, in the first implementation of the 6th aspect, the processing module is specifically used for：Read institute
State each data corresponding to address in specified address；
The sending module is specifically used for：Number corresponding to each address in the specified address is sent to the first end
According to.
With reference to the 6th aspect, in second of implementation of the 6th aspect, exist in second end multiple scattered interior
Region is deposited, the specified address includes the multiple addresses being distributed in the multiple scattered region of memory, then
The processing module is specifically used for：Read each data corresponding to address in the specified address；
The sending module is specifically used for：Number corresponding to each address in the specified address is sent to the first end
According to.
With reference to the 6th aspect, in the third implementation of the 6th aspect, described device also includes：
Registering modules, before the RDMA Read Request message that the receiving module receives that first end is sent, for certainly
The region of memory of body is registered to network interface card.
7th aspect, there is provided the device that data write under a kind of RDMA, described device include：
Processing module, for by need write data specified address and the specified address corresponding to data be encapsulated in
RDMA Write enters in data message, and the specified address is multiple addresses scattered in the range of the region of memory at second end；
Sending module, enter data message for sending the RDMA Write to second end.
With reference to the 7th aspect, in the first implementation of the 7th aspect, described device also includes：
Acquisition module, in the processing module that the specified address and the specified address that need write-in data is corresponding
Data be encapsulated in before RDMA Write enters in data message, obtain the memory information of the region of memory at second end, wherein, institute
Stating the memory information of region of memory includes：The address information and/or length information of region of memory.
With reference to the first implementation of the 7th aspect, in second of implementation of the 7th aspect, second end
Region of memory include multiple scattered region of memorys, the acquisition module is specifically used for：
Obtain the memory information of each scattered region of memory in the region of memory at second end.
With reference to the 7th aspect second of implementation, the 7th aspect the third implementation in, it is described specifiedly
Location includes the multiple addresses being distributed in the multiple scattered region of memory.
With reference to the 7th aspect, in the 4th kind of implementation of the 7th aspect, described device also includes：
Receiving module, after entering data message to second end transmission RDMA Write in the sending module,
Receive the confirmation message that the instruction that second end is sent is successfully written.
Eighth aspect, there is provided the device that data write under a kind of RDMA, described device include：
Receiving module, the RDMA Write for receiving first end transmission enter data message, and the RDMA Write enters in data message
Data corresponding to the specified address for needing to write data and the specified address are carried, the specified address is second end
The multiple addresses disperseed in the range of region of memory；
Processing module, for writing data corresponding to the specified address in the specified address.
With reference to eighth aspect, in the first implementation of eighth aspect, exist in second end multiple scattered interior
Region is deposited, the specified address includes the multiple addresses being distributed in the multiple scattered region of memory,
The processing module is specifically used for：Each address in the specified address writes number corresponding to the address
According to.
9th aspect, there is provided a kind of network system, the network system are any including being mentioned in the above-mentioned 5th aspect
The device for any RDMA digital independents mentioned in the device of RDMA digital independents and above-mentioned 6th aspect.
Tenth aspect, there is provided a kind of network system, the network system are any including being mentioned in the above-mentioned 7th aspect
The device for any RDMA data write-in mentioned in the device of RDMA data write-in and above-mentioned eighth aspect.
After adopting the above technical scheme, under RDMA provided in an embodiment of the present invention reading and writing data method, apparatus and network
System, first end will can need to read the more of data when carrying out RDMA read-write operations in the range of the region of memory at the second end
Individual scattered address is sent to the second end, and the read-write of the data to disperseing address is performed by the second end, realizes reading and writing data under RDMA
Operation.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, make required in being described below to embodiment
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for
For those of ordinary skill in the art, on the premise of not paying creative work, other can also be obtained according to these accompanying drawings
Accompanying drawing.
Fig. 1 is the flow chart of the method for digital independent under a kind of RDMA provided in an embodiment of the present invention；
Fig. 2 is the flow chart of the method for digital independent under another RDMA provided in an embodiment of the present invention；
Fig. 3 is the flow chart of the method that data write under a kind of RDMA provided in an embodiment of the present invention；
Fig. 4 is the flow chart of the method that data write under another RDMA provided in an embodiment of the present invention；
Fig. 5 is the schematic diagram of the method for digital independent under the RDMA that one embodiment of the invention provides；
Fig. 6 is the schematic diagram of the method that data write under the RDMA that one embodiment of the invention provides；
Fig. 7a-7c is the structured flowchart of the device of digital independent under RDMA provided in an embodiment of the present invention；
Fig. 8 a-8b are the structured flowcharts of the device of digital independent under RD MA provided in an embodiment of the present invention；
Fig. 9 a-9c are the structured flowcharts of the device that data write under RDMA provided in an embodiment of the present invention；
Figure 10 is the structured flowchart of the device that data write under RDMA provided in an embodiment of the present invention.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention
Formula is described in further detail.
Fig. 1 is the flow chart of the method for digital independent under a kind of RDMA provided in an embodiment of the present invention, reference picture 1, this hair
The method of digital independent includes under the RDMA that bright embodiment provides：
11st, first end sends RDMA Read Request message to the second end, and being carried in the RDMA Read Request message needs
The specified address of data is read, the specified address is multiple addresses scattered in the range of the region of memory at second end.
It is pointed out that they are made a distinction by first end and the status equity at the second end in the embodiment of the present invention
Only to facilitate description.In embodiments of the present invention, the first end can be client, and second end can be service
End；Conversely, the first end can also be service end, second end can be client.
Alternatively, before the first end of this step 11 sends RDMA Read Request message to the second end, the embodiment of the present invention
The method of digital independent may also include under the RDMA of offer：The first end obtains the internal memory letter of the region of memory at second end
Breath.Wherein, the memory information of the region of memory may include：The address information and/or length information of region of memory.
Further, when the region of memory at second end includes multiple scattered region of memorys, the first end obtains
The memory information of the region of memory at second end may include：
The first end obtains the memory information of each scattered region of memory in the region of memory at second end.
Wherein, the memory information of region of memory of the present invention to specifically how to obtain second end is not construed as limiting, any
First end with can be used by way of the memory information for interacting the region of memory for obtaining second end at the second end.
Meanwhile before first end described in this step sends RDMA Read Request message to the second end, the first end is also
The region of memory of itself can be registered to network interface card.
12nd, the first end receives data corresponding to specified address that second end is sent, described.
Wherein, first end described in this step receive specified address that second end is sent, described corresponding to data can wrap
Include：Data corresponding to each address in specified address that first end reception second end is sent, described.
When multiple scattered region of memorys be present in second end, each address in the specified address may be distinguished
It is present in different scattered region of memorys, now, data corresponding to each address in the specified address come from each
Data in the scattered region of memory of difference.
In addition, first end to the RDMA Read Request message that the second end is sent in addition to it can carry the specified address, also
The memory information for disperseing region of memory in the specified address corresponding to each address can be carried.So, the second end is being known
After disperseing the memory information of region of memory corresponding to each address in the specified address, you can in corresponding scattered region of memory
Read data corresponding to the address.
The method of digital independent under RDMA provided in an embodiment of the present invention, first end is when carrying out RDMA read operations, energy
It is enough that the multiple scattered addresses for needing to read data in the range of the region of memory at second end are sent to the second end, by the second end
The data of each scattered address are read, realize data read operation under RDMA.
Fig. 2 is the flow chart of the method for digital independent under another RDMA provided in an embodiment of the present invention, reference picture 2, this
The method that data write under the RDMA that inventive embodiments provide includes：
21st, the second end receives the RDMA Read Request message that first end is sent, and is carried in the RDMA Read Request message
Need read data specified address, the specified address for second end region of memory in the range of disperse multiplely
Location.
Wherein, alternatively, also carried in the RDMA Read Request message in the specified address corresponding to each address
The memory information of scattered region of memory.The memory information of the scattered region of memory may include：The address letter of scattered region of memory
Breath and/or length information.
Meanwhile before the RDMA Read Request message that the second end described in step 21 receives that first end is sent, described first
End can also register to the region of memory of itself to network interface card.
22nd, data corresponding to the specified address are read at second end.
Wherein, when scattered multiple addresses are present in the same region of memory in second end, described second
Data corresponding to the reading specified address of end may include：Read in the specified address corresponding to each address at second end
Data.
Multiple scattered region of memorys in second end be present, the specified address includes being distributed in the multiple disperse
During multiple addresses in region of memory, step 22 can be specially：Read each address pair in the specified address in second end
The data answered, wherein data corresponding to each address can be located in different scattered region of memorys.
Also carried in the RDMA Read Request message and disperse memory field in the specified address corresponding to each address
During the memory information in domain, the second end directly can disperse in region of memory to read number corresponding to the address corresponding to each address
According to.
23rd, second end sends data corresponding to the specified address to the first end.
Wherein, data corresponding to the second end described in step 23 to the first end transmission specified address may include：Institute
State the second end and send data corresponding to each address in the specified address to the first end.
Multiple scattered region of memorys in second end be present, the specified address includes being distributed in the multiple disperse
During multiple addresses in region of memory, this step 23 can be specially：Second end sends described specified to the first end
Each data corresponding to address in location.Wherein, data corresponding to each address can be located in different scattered region of memorys.
The method of digital independent under RDMA provided in an embodiment of the present invention, first end are obtaining the scattered memory field at the second end
After the memory information in domain, when carrying out RDMA read operations, it will can need to read in the range of the scattered region of memory at second end
The address of access evidence is sent to the second end, and the data of the address are read by the second end, realize data read operation under RDMA.
Fig. 3 be under a kind of RDMA provided in an embodiment of the present invention data write method flow chart, reference picture 3, this hair
The method that data write under the RDMA that bright embodiment provides includes：
31st, data corresponding to the specified address for needing write-in data and the specified address are encapsulated in RDMA Write by first end
Enter in data message, the specified address is multiple addresses scattered in the range of the region of memory at second end.
Wherein, will need to write corresponding to specified address and the specified address of data in first end described in this step 31
Data are encapsulated in before RDMA Write enters in data message, and the first end can obtain the internal memory letter of the region of memory at second end
Breath.Wherein, the memory information of the region of memory may include：The address information and/or length information of region of memory.
Further, the region of memory at second end may include multiple scattered region of memorys, and now, the first end obtains
Taking the memory information of the region of memory at second end may include：
The first end obtains the memory information of each scattered region of memory in the region of memory at second end.
The memory information of region of memory of the present invention to specifically how to obtain second end is not construed as limiting, any by the
One end can use with the mode of the memory information for interacting the region of memory for obtaining second end at the second end.
Alternatively, when multiple scattered region of memorys be present in second end, the specified address may include to be distributed in
Multiple addresses in the multiple scattered region of memory.
In embodiments of the present invention, the RDMA Write, which enters, can also carry each address in the specified address in data message
The memory information of corresponding scattered region of memory.
Meanwhile before this step 31, first end can also be registered to the region of memory of itself to network interface card.
32nd, the first end enters data message to second end transmission RDMA Write.
After first end described in this step enters data message to second end transmission RDMA Write, the first end
The confirmation message that the instruction that second end is sent is successfully written can be received.
The method that data write under RDMA provided in an embodiment of the present invention, first end is when carrying out RDMA Write operation, energy
Enough multiple scattered addresses that will need to write data in the range of the region of memory at second end and corresponding data are sent to the
Two ends, the corresponding data in the address are write in each address by the second end, realize data write operation under RDMA.
Fig. 4 be under a kind of RDMA provided in an embodiment of the present invention data write method flow chart, reference picture 4, this hair
The method that data write under the RDMA that bright embodiment provides includes：
41st, the second end, which receives first end and builds the RDMA Write of transmission, enters data message, and the RDMA Write, which enters in data message, to be taken
Band needs to write data corresponding to the specified address of data and the specified address, and the specified address is interior for second end
Deposit the multiple addresses disperseed in regional extent.
Wherein, received in the second end described in step 41 first end build transmission RDMA Write enter data message before, described the
Two ends can register to the region of memory of itself.
42nd, second end writes data corresponding to the specified address in the specified address.
Wherein, multiple scattered region of memorys in second end be present, the specified address is described more including being distributed in
During multiple addresses in individual scattered region of memory, the second end described in step 42 can in data corresponding to the specified address write-in
Including：Each address of second end in the specified address writes data corresponding to the address.Wherein, each address
Corresponding data can be located in different scattered region of memorys
In embodiments of the present invention, the RDMA Write, which enters, can also carry each address in the specified address in data message
The memory information of corresponding scattered region of memory.Wherein, the memory information of the scattered region of memory may include：Scattered memory field
The address information and/or length information in domain.
The method that data write under RDMA provided in an embodiment of the present invention, first end is when carrying out RDMA Write operation, energy
Enough multiple scattered addresses that will need to write data in the range of the region of memory at second end and corresponding data are sent to the
Two ends, the corresponding data in the address are write in each address by the second end, realize data write operation under RDMA.
Technical scheme for a better understanding of the present invention, scattered internal memory is read or write below to service end with client
Exemplified by address, by specific embodiment, the present invention will be further elaborated.
Fig. 5 is the schematic diagram of the method for digital independent under the RDMA that one embodiment of the invention provides.Reference picture 5, the present invention
The method of digital independent includes under the RDMA that embodiment provides：
51st, client specifies addr parameters, registers memory region to network interface card by ibv_reg_mr interfaces；Service end is adopted
With the same manner registers memory region to network interface card.
52nd, the client obtains the memory information of the region of memory of the service end.Wherein, the internal memory of the service end
Region, which includes scattered region of memory and contiguous memory region, the memory information of the region of memory, to be included：The address of region of memory
Information and/or length information.In the embodiment of the present invention, client can obtain the internal memory of service end by a variety of modes
The memory information in region, include but is not limited to, it is described during the client and service end establish RDMA communication connections
Client obtains the memory information of the region of memory in the service end by privately owned passage.
53rd, client sends RDMA Read Request message to service end, and being carried in the RDMA Read Request message needs
The specified address of data is read, wherein, the specified address is multiple scattered ground in the range of the region of memory of the service end
Location.Alternatively, if service end includes multiple scattered region of memorys, the specified address may include to be distributed in the multiple point
Dissipate multiple addresses in region of memory.The RDMA Read Request message, which may also include, disperses internal memory corresponding to the specified address
The memory information in region.
54th, service end receives the RDMA Read Request message that client is sent, and reads data corresponding to the specified address,
And send data corresponding to the specified address to the client.Specifically, the specified address can be read in the service end
In each address corresponding to data, and send to client data corresponding to each address in the specified address.
Wherein, in embodiments of the present invention, if multiple scattered region of memorys in the service end be present, the specified address
It may include the multiple addresses being distributed in the multiple scattered region of memory.
If multiple scattered region of memorys in the service end be present, and the specified address includes being distributed in the multiple point
Multiple addresses in region of memory are dissipated, data corresponding to the service end reading specified address can be specially：The service
Read data corresponding to each address in the specified address in end.Meanwhile the service end sends the finger to the client
Determining data corresponding to address can be specially：The service end sends each address in the specified address to the client and corresponded to
Data.
So that client reads three scattered memory address to service end as an example, the protocol header of RDMA Read Request message can be such as
Under：
That is, each SrcTO both corresponds to a scattered memory address.Three SrcTO be present in three scattered memory address.
Wherein, each field meanings are as follows:
SinkSTag:Data receiver memory headroom identifies；
SinkTO:The offset of data initial address when receiving terminal memory headroom stores；
RDMARDSZ：RDMA Read message size；
SrcSTag:Data source end memory space identification；
SrcTO:The offset of data initial address when source memory headroom is read.
From the foregoing, it will be observed that by any two-server of network connection, agreement is read by changing RDMA in rdma protocol
The capability simultaneously can once will by head field to the network interface card with RDMA abilities, client by way of RDMA readings
The digital independent of service end even scattered memory address is to local terminal memory headroom.
Fig. 6 is the schematic diagram of the method that data write under the RDMA that one embodiment of the invention provides.Reference picture 6, the present invention
The method of digital independent includes under the RDMA that embodiment provides：
61st, client specifies addr parameters, registers memory region to network interface card by ibv_reg_mr interfaces；Service end is adopted
With the same manner registers memory region to network interface card.
62nd, the client obtains the memory information of the region of memory of the service end.Wherein, the internal memory of the service end
Region, which includes scattered region of memory and contiguous memory region, the memory information of the region of memory, to be included：The address of region of memory
Information and/or length information.In the embodiment of the present invention, client can obtain the internal memory of service end by a variety of modes
The memory information in region, include but is not limited to, it is described during the client and service end establish RDMA communication connections
Client can obtain the memory information of the scattered region of memory in the service end by privately owned passage.
63rd, data corresponding to specified address, the specified address are encapsulated in RDMA Write and enter data message by the client
In, and send the RDMA Write to the service end and enter data message.Wherein, the specified address is the internal memory of the service end
Multiple scattered addresses of write-in data are needed in regional extent.Alternatively, if service end includes multiple scattered region of memorys,
The specified address may include the multiple addresses being distributed in the multiple scattered region of memory.The RDMA Write enters data message
It may also include the memory information for disperseing region of memory corresponding to the specified address.
64th, the RDMA Write that the service end reception client is sent enters data message, and is write in the specified address
Enter data corresponding to the specified address.
Wherein, in embodiments of the present invention, if multiple scattered region of memorys in the service end be present, the specified address
It may include the multiple addresses being distributed in the multiple scattered region of memory.
The service end can be specially in data corresponding to the specified address write-in：The service end is described specified
Each address in location writes data corresponding to the address.
So that client writes three scattered memory address to service end as an example, the protocol header that RDMA Write enters data message can be such as
Under：
That is, each Data Sink Tagged offset correspond to a scattered memory address.Three scattered memory address
Three Data Sink Tagged offset be present.
Wherein, each field meanings are as follows：
Data Sink STag:Data receiver memory headroom identifies；
Data Sink Tagged offset:The offset of data initial address when receiving terminal memory headroom stores；
RDMA write ULP Payload:Data payload.
From the foregoing, it will be observed that by any two-server of network connection, enter agreement by changing RDMA Write in rdma protocol
The capability simultaneously can once will by head field to the network interface card with RDMA abilities, client by way of RDMA Write enters
Client disperses the data write service end memory space of memory address.
The method of reading and writing data has following excellent under RDMA provided by the invention it can be seen from above example description
Point：
First, traditional RDMA read operations do not support by service end disperse memory address data read in client it is continuous or
Scattered memory address, only support service end contiguous memory address reading in client memory address, the present invention supports reading service
The scattered memory address in end and compatible tradition read the function of contiguous memory address.
Second, traditional RDMA Write operation is not supported client internal storage data write service end disperseing memory address,
Write service end contiguous memory address is only supported, the present invention supports write service end to disperse memory address and compatible tradition write-in clothes
The function of business end contiguous memory address.
In addition, it is desirable to explanation, the scene described of the invention supports virtual machine environment, physical machine environment, Linux
Any platform such as system, Windows systems and operating system.
It should be noted that comparatively " service end " and " client " in the embodiment of the present invention is, it is right in the text
They are made a distinction merely to convenient on describing, and are not to mean limitation, for example, in embodiments of the present invention, they can be with
Used interchangeably, i.e. service end can be client, and client can also be service end.
Accordingly, the embodiment of the present invention also provides a kind of device of digital independent under RDMA.Reference picture 7a, the present invention are implemented
The device 70 of digital independent includes sending module 71 and receiving module 72 under the RDMA that example provides.Wherein：
Sending module 71, for sending RDMA Read Request message to the second end, taken in the RDMA Read Request message
Band need read data specified address, the specified address for second end region of memory in the range of disperse multiplely
Location；
Receiving module 72, for receiving data corresponding to specified address that second end is sent, described.
Alternatively, reference picture 7b, in one embodiment, described device 70 also include acquisition module 73.Wherein：
Acquisition module 73, for before the sending module 71 sends RDMA Read Request message to the second end, obtaining
The memory information of the region of memory at second end, wherein, the memory information includes：Address information and/or length information.
Further, when the region of memory at second end includes multiple scattered region of memorys, the acquisition module 73
It can be specifically used for：
Obtain the memory information of each scattered region of memory in the region of memory at second end.
Further, the multiple addresses being distributed in the multiple scattered region of memory are included in the specified address
When, the receiving module 72 can be specifically used for：
Receive data corresponding to each address in specified address that second end is sent, described.
Alternatively, reference picture 7c, in one embodiment, described device also include：Registering modules 74, in the hair
Send module to the second end send RDMA Read Request message before, the region of memory of itself is registered to network interface card.
The device of digital independent under RDMA provided in an embodiment of the present invention, first end is when carrying out RDMA read operations, energy
It is enough that the multiple scattered addresses for needing to read data in the range of the region of memory at second end are sent to the second end, by the second end
The data of each scattered address are read, realize data read operation under RDMA.
Fig. 8 a are the structured flowcharts of the device of another RDMA digital independents provided in an embodiment of the present invention.Reference picture 8a,
The device 80 of digital independent includes receiving module 81, processing module 82 and sending module under RDMA provided in an embodiment of the present invention
83.Wherein：
Receiving module 81, for receiving the RDMA Read Request message of first end transmission, the RDMA Read Request message
Middle carrying needs to read the specified address of data, and the specified address is more to disperse in the range of the region of memory at second end
Individual address；
Processing module 82, for reading data corresponding to the specified address；
Sending module 83, for sending data corresponding to the specified address to the first end.
Alternatively, in one embodiment, the processing module 82 can be specifically used for：Read each in the specified address
Data corresponding to address.The sending module 83 can be specifically used for：Sent to the first end in the specified address each
Data corresponding to location.
Alternatively, when multiple scattered region of memorys be present in second end, the specified address includes being distributed in institute
Multiple addresses in multiple scattered region of memorys are stated, then
The processing module 82 is specifically used for：Read each data corresponding to address in the specified address；
The sending module 83 is specifically used for：Number corresponding to each address in the specified address is sent to the first end
According to.
Alternatively, reference picture 8b, in one embodiment, described device 80 also include：
Registering modules 84, before the RDMA Read Request message that the receiving module 81 receives that first end is sent, it is used for
The region of memory of itself is registered to network interface card.
The device of digital independent under RDMA provided in an embodiment of the present invention, first end are obtaining the scattered memory field at the second end
After the memory information in domain, when carrying out RDMA read operations, it will can need to read in the range of the scattered region of memory at second end
The address of access evidence is sent to the second end, and the data of the address are read by the second end, realize data read operation under RDMA.
Fig. 9 a are the structured flowcharts of the device of another RDMA data write-in provided in an embodiment of the present invention.Reference picture 9a,
The device 90 that data write under RDMA provided in an embodiment of the present invention includes processing module 91 and sending module 92.Wherein：
Processing module 91, for by need write data specified address and the specified address corresponding to data be encapsulated in
RDMA Write enters in data message, and the specified address is multiple addresses scattered in the range of the region of memory at second end；
Sending module 92, enter data message for sending the RDMA Write to second end.
Alternatively, reference picture 9b, in one embodiment, described device 90 also include：
Acquisition module 93, for will need to write the specified address of data and the specified address in the processing module 91
Corresponding data are encapsulated in before RDMA Write enters in data message, obtain the memory information of the region of memory at second end, its
In, the memory information of the region of memory includes：The address information and/or length information of region of memory.
Further, when the region of memory at second end includes multiple scattered region of memorys, the acquisition module 93
It can be specifically used for：Obtain the memory information of each scattered region of memory in the region of memory at second end.
Alternatively, the specified address in the embodiment of the present invention may include to be distributed in the multiple scattered region of memory
Multiple addresses.
Alternatively, reference picture 9c, in one embodiment, described device 90 also include：
Receiving module 94, for the sending module to second end send the RDMA Write enter data message it
Afterwards, the confirmation message that the instruction that second end is sent is successfully written is received.
The device that data write under RDMA provided in an embodiment of the present invention, first end is when carrying out RDMA Write operation, energy
Enough multiple scattered addresses that will need to write data in the range of the region of memory at second end and corresponding data are sent to the
Two ends, the corresponding data in the address are write in each address by the second end, realize data write operation under RDMA.
Figure 10 is the structured flowchart of the device of another RDMA data write-in provided in an embodiment of the present invention.Reference picture 10,
The device 100 that data write under RDMA provided in an embodiment of the present invention includes receiving module 101 and processing module 102.Wherein：
Receiving module 101, the RDMA Write for receiving first end transmission enter data message, and the RDMA Write enters data message
Middle carrying needs to write data corresponding to the specified address of data and the specified address, and the specified address is second end
Region of memory in the range of disperse multiple addresses；
Processing module 102, for writing data corresponding to the specified address in the specified address.
Alternatively, when multiple scattered region of memorys be present in second end, the specified address may include to be distributed in
Multiple addresses in the multiple scattered region of memory.Now, the processing module 102 is specifically used for：In the specified address
In each address write data corresponding to the address.
The method that data write under RDMA provided in an embodiment of the present invention, first end is when carrying out RDMA Write operation, energy
Enough multiple scattered addresses that will need to write data in the range of the region of memory at second end and corresponding data are sent to the
Two ends, the corresponding data in the address are write in each address by the second end, realize data write operation under RDMA.
In addition, the embodiment of the present invention also provides a kind of network system, it may include shown in any figure in Fig. 7 a- Fig. 7 c
RDMA reading devices and the RDMA reading devices shown in Fig. 8 a or Fig. 8 b.
In addition, the embodiment of the present invention also provides another network system, it may include shown in any figure in Fig. 9 a- Fig. 9 c
RDMA Write enter device and the RDMA Write shown in Figure 10 enters device.
It should be noted that：The dress that data write under the device and RDMA of digital independent under the RDMA that above-described embodiment provides
Put, only with the division progress of above-mentioned each functional module for example, in practical application, as needed can divide above-mentioned function
With by different functional module completions, i.e., the internal structure of device is divided into different functional modules, to complete above description
All or part of function.In addition, the device of reading and writing data and RDMA reading and writing datas under the RDMA that above-described embodiment provides
Embodiment of the method belongs to same design, and its specific implementation process refers to embodiment of the method, repeats no more here.
One of ordinary skill in the art will appreciate that hardware can be passed through by realizing all or part of step of above-described embodiment
To complete, by program the hardware of correlation can also be instructed to complete, described program can be stored in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only storage, disk or CD etc..
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent substitution and improvements made etc., it should be included in the scope of the protection.
Claims (35)
1. a kind of method that Remote Direct Memory accesses digital independent under RDMA, it is characterised in that methods described includes：
First end sends RDMA Read Request message to the second end, and being carried in the RDMA Read Request message needs to read data
Specified address, the specified address for second end region of memory in the range of disperse multiple addresses, it is described scattered
Multiple Address d istributions are in multiple scattered region of memorys；
The first end receives data corresponding to specified address that second end is sent, described.
2. according to the method for claim 1, it is characterised in that send RDMA Read Request to the second end in the first end
Before message, methods described also includes：
The first end obtains the memory information of the region of memory at second end, wherein, the memory information includes：Believe address
Breath and/or length information.
3. according to the method for claim 2, it is characterised in that the region of memory at second end includes multiple scattered internal memories
Region, the memory information that the first end obtains the region of memory at second end include：
The first end obtains the memory information of each scattered region of memory in the region of memory at second end.
4. according to the method for claim 3, it is characterised in that the specified address includes being distributed in the multiple disperse
Multiple addresses in region are deposited, then
Data corresponding to specified address that first end reception second end is sent, described include：
Data corresponding to each address in specified address that first end reception second end is sent, described.
5. according to the method for claim 1, it is characterised in that send RDMA Read Request to the second end in the first end
Before message, methods described also includes：
The first end is registered to the region of memory of itself to network interface card.
6. a kind of method that Remote Direct Memory accesses digital independent under RDMA, it is characterised in that methods described includes：
Second end receives the RDMA Read Request message that first end is sent, and being carried in the RDMA Read Request message needs to read
The specified address of data, the specified address for second end region of memory in the range of disperse multiple addresses, described point
Scattered multiple Address d istributions are in multiple scattered region of memorys；
Read data corresponding to the specified address in second end；
Second end sends data corresponding to the specified address to the first end.
7. according to the method for claim 6, it is characterised in that read data corresponding to the specified address in second end
Including：Read data corresponding to each address in the specified address in second end；
Data corresponding to second end to the first end transmission specified address include：Second end is to described first
End sends data corresponding to each address in the specified address.
8. according to the method for claim 6, it is characterised in that multiple scattered region of memorys, institute in second end be present
Multiple addresses that specified address includes being distributed in the multiple scattered region of memory are stated, then
Data corresponding to the second end reading specified address include：Read each in the specified address in second end
Data corresponding to address；
Data corresponding to second end to the first end transmission specified address include：Second end is to described first
End sends data corresponding to each address in the specified address.
9. according to the method for claim 6, it is characterised in that the RDMA that first end is sent is received at second end and is read
Before request message, methods described also includes：
The region of memory of the two port itself is registered to network interface card.
10. a kind of Remote Direct Memory accesses the method that data write under RDMA, it is characterised in that methods described includes：
Data corresponding to the specified address for needing write-in data and the specified address are encapsulated in RDMA Write and enter data by first end
In message, the specified address for the second end region of memory in the range of disperse multiple addresses, scattered multiple addresses
It is distributed in multiple scattered region of memorys；
The first end sends the RDMA Write to second end and enters data message.
11. according to the method for claim 10, it is characterised in that will need to write the specified of data in the first end
Data corresponding to location and the specified address are encapsulated in before RDMA Write enters in data message, and methods described also includes：
The first end obtains the memory information of the region of memory at second end, wherein, the memory information of the region of memory
Including：The address information and/or length information of region of memory.
12. according to the method for claim 11, it is characterised in that the region of memory at second end includes multiple scattered interior
Region is deposited, the memory information that the first end obtains the region of memory at second end includes：
The first end obtains the memory information of each scattered region of memory in the region of memory at second end.
13. according to the method for claim 12, it is characterised in that the specified address includes being distributed in the multiple point
Dissipate multiple addresses in region of memory.
14. according to the method for claim 10, it is characterised in that described in the first end to second end transmission
After RDMA Write enters data message, methods described also includes：
The first end receives the confirmation message that the instruction that second end is sent is successfully written.
15. a kind of Remote Direct Memory accesses the method that data write under RDMA, it is characterised in that methods described includes：
Second end receives the RDMA Write that first end is sent and enters data message, and the RDMA Write, which enters to carry in data message, to be needed to write
Data corresponding to the specified address of data and the specified address, the specified address are the region of memory scope at second end
Interior scattered multiple addresses, scattered multiple Address d istributions are in multiple scattered region of memorys；
Second end writes data corresponding to the specified address in the specified address.
16. according to the method for claim 15, it is characterised in that multiple scattered region of memorys in second end be present,
The specified address includes the multiple addresses being distributed in the multiple scattered region of memory,
Second end includes in data corresponding to the specified address write-in：Second end is every in the specified address
One address writes data corresponding to the address.
17. a kind of Remote Direct Memory accesses the device of digital independent under RDMA, it is characterised in that described device includes：
Sending module, for sending RDMA Read Request message to the second end, being carried in the RDMA Read Request message needs
Read data specified address, the specified address for second end region of memory in the range of disperse multiple addresses, institute
Scattered multiple Address d istributions are stated in multiple scattered region of memorys；
Receiving module, for receiving data corresponding to specified address that second end is sent, described.
18. device according to claim 17, it is characterised in that described device also includes：
Acquisition module, for before the sending module sends RDMA Read Request message to the second end, obtaining described second
The memory information of the region of memory at end, wherein, the memory information includes：Address information and/or length information.
19. device according to claim 18, it is characterised in that the region of memory at second end includes multiple scattered interior
Region is deposited, then the acquisition module is specifically used for：
Obtain the memory information of each scattered region of memory in the region of memory at second end.
20. device according to claim 19, it is characterised in that the specified address includes being distributed in the multiple point
Multiple addresses in region of memory are dissipated, then
The receiving module is specifically used for：
Receive data corresponding to each address in specified address that second end is sent, described.
21. device according to claim 17, it is characterised in that described device also includes：
Registering modules, for the sending module to the second end send RDMA Read Request message before, to the internal memory of itself
Registered to network interface card in region.
22. a kind of Remote Direct Memory accesses the device of digital independent under RDMA, it is characterised in that described device includes：
Receiving module, for receiving the RDMA Read Request message of first end transmission, carried in the RDMA Read Request message
Need read data specified address, the specified address for the second end region of memory in the range of disperse multiple addresses, institute
Scattered multiple Address d istributions are stated in multiple scattered region of memorys；
Processing module, for reading data corresponding to the specified address；
Sending module, for sending data corresponding to the specified address to the first end.
23. device according to claim 22, it is characterised in that the processing module is specifically used for：Read described specify
Each data corresponding to address in address；
The sending module is specifically used for：Data corresponding to each address in the specified address are sent to the first end.
24. device according to claim 22, it is characterised in that multiple scattered region of memorys in second end be present,
The specified address includes the multiple addresses being distributed in the multiple scattered region of memory, then
The processing module is specifically used for：Read each data corresponding to address in the specified address；
The sending module is specifically used for：Data corresponding to each address in the specified address are sent to the first end.
25. device according to claim 22, it is characterised in that described device also includes：
Registering modules, the receiving module receive first end send RDMA Read Request message before, for itself
Region of memory is registered to network interface card.
26. a kind of Remote Direct Memory accesses the device that data write under RDMA, it is characterised in that described device includes：
Processing module, for by need write data specified address and the specified address corresponding to data be encapsulated in RDMA Write
Enter in data message, the specified address is multiple addresses for disperseing in the range of the region of memory at the second end, described scattered more
Individual Address d istribution is in multiple scattered region of memorys；
Sending module, enter data message for sending the RDMA Write to second end.
27. device according to claim 26, it is characterised in that described device also includes：
Acquisition module, for will be counted in the processing module corresponding to the specified address for needing write-in data and the specified address
According to being encapsulated in before RDMA Write enters in data message, obtain the memory information of the region of memory at second end, wherein, it is described in
Depositing the memory information in region includes：The address information and/or length information of region of memory.
28. device according to claim 27, it is characterised in that the region of memory at second end includes multiple scattered interior
Region is deposited, the acquisition module is specifically used for：
Obtain the memory information of each scattered region of memory in the region of memory at second end.
29. device according to claim 28, it is characterised in that the specified address includes being distributed in the multiple disperse
Multiple addresses in region of memory.
30. device according to claim 26, it is characterised in that described device also includes：
Receiving module, after entering data message to second end transmission RDMA Write in the sending module, receive
The confirmation message that the instruction that second end is sent is successfully written.
31. a kind of Remote Direct Memory accesses the device that data write under RDMA, it is characterised in that described device includes：
Receiving module, the RDMA Write for receiving first end transmission enter data message, and the RDMA Write, which enters in data message, to be carried
Need write data specified address and the specified address corresponding to data, the specified address be the second end region of memory
In the range of disperse multiple addresses, scattered multiple Address d istributions are in multiple scattered region of memorys；
Processing module, for writing data corresponding to the specified address in the specified address.
32. device according to claim 31, it is characterised in that multiple scattered region of memorys in second end be present,
The specified address includes the multiple addresses being distributed in the multiple scattered region of memory, and the processing module is specifically used for：
Each address in the specified address writes data corresponding to the address.
33. a kind of network system, it is characterised in that including the Remote Direct Memory any one of claim 17-21
Access the device of digital independent under the RDMA under RDMA any one of the device and claim 22-25 of digital independent.
34. a kind of network system, it is characterised in that including the Remote Direct Memory any one of claim 26-30
Access the device that data write under the RDMA any one of the device and claim 31-32 that data write under RDMA.
35. a kind of computer-readable recording medium, it is characterised in that the storage medium internal memory contains computer program, described
The step in the method described in claim any one of 1-16 is realized when computer program is executed by processor.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410108028.7A CN103929415B (en) | 2014-03-21 | 2014-03-21 | The method, apparatus and network system of reading and writing data under RDMA |
| PCT/CN2014/091263 WO2015139470A1 (en) | 2014-03-21 | 2014-11-17 | Data write/read method and apparatus in rdma and network system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410108028.7A CN103929415B (en) | 2014-03-21 | 2014-03-21 | The method, apparatus and network system of reading and writing data under RDMA |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103929415A CN103929415A (en) | 2014-07-16 |
| CN103929415B true CN103929415B (en) | 2018-03-06 |
Family
ID=51147493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410108028.7A Active CN103929415B (en) | 2014-03-21 | 2014-03-21 | The method, apparatus and network system of reading and writing data under RDMA |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN103929415B (en) |
| WO (1) | WO2015139470A1 (en) |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103929415B (en) * | 2014-03-21 | 2018-03-06 | 华为技术有限公司 | The method, apparatus and network system of reading and writing data under RDMA |
| CN105472023B (en) * | 2014-12-31 | 2018-11-20 | 华为技术有限公司 | A kind of method and device of direct distance input and output |
| CN105487937A (en) * | 2015-11-27 | 2016-04-13 | 华为技术有限公司 | RDMA (Remote Direct Memory Access) implementation method and device |
| CN105630426A (en) * | 2016-01-07 | 2016-06-01 | 清华大学 | Method and system for obtaining remote data based on RDMA (Remote Direct Memory Access) characteristics |
| US10769098B2 (en) * | 2016-04-04 | 2020-09-08 | Marvell Asia Pte, Ltd. | Methods and systems for accessing host memory through non-volatile memory over fabric bridging with direct target access |
| CN105893323A (en) * | 2016-05-23 | 2016-08-24 | 华为技术有限公司 | Data reading method and data reading equipment |
| CN106487896B (en) * | 2016-10-14 | 2019-10-08 | 北京百度网讯科技有限公司 | Method and apparatus for handling remote direct memory access request |
| CN106873915A (en) * | 2017-02-22 | 2017-06-20 | 郑州云海信息技术有限公司 | A kind of data transmission method and device based on RDMA registers memory blocks |
| CN107547623B (en) * | 2017-06-28 | 2021-02-05 | 新华三云计算技术有限公司 | Application data transmission method and device, application server and application proxy server |
| CN107493329A (en) * | 2017-08-15 | 2017-12-19 | 诸葛晴凤 | A kind of remote document access method of distributed memory file system |
| CN108228476B (en) * | 2017-12-22 | 2021-02-09 | 新华三技术有限公司 | Data acquisition method and device |
| CN110535811B (en) * | 2018-05-25 | 2022-03-04 | 中兴通讯股份有限公司 | Remote memory management method and system, server, client and storage medium |
| CN111858418B (en) * | 2019-04-30 | 2023-04-07 | 华为技术有限公司 | Memory communication method and device based on remote direct memory access RDMA |
| CN110177118B (en) * | 2019-06-13 | 2022-03-29 | 上海海事大学 | RDMA-based RPC communication method |
| CN110191194B (en) * | 2019-06-13 | 2020-07-03 | 华中科技大学 | RDMA (remote direct memory Access) network-based distributed file system data transmission method and system |
| WO2022061763A1 (en) * | 2020-09-25 | 2022-03-31 | 华为技术有限公司 | Data storage method and apparatus |
| CN112954068B (en) * | 2021-03-09 | 2022-09-27 | 西安奥卡云数据科技有限公司 | RDMA (remote direct memory Access) -based data transmission method and device |
| WO2022218160A1 (en) * | 2021-04-14 | 2022-10-20 | 华为技术有限公司 | Data access system and method, and device and network card |
| CN113377704B (en) * | 2021-06-11 | 2022-04-12 | 上海壁仞智能科技有限公司 | Artificial intelligence chip and data operation method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1930563A (en) * | 2004-01-22 | 2007-03-14 | 高通股份有限公司 | A multiple address two channel bus structure |
| CN101459676A (en) * | 2008-12-31 | 2009-06-17 | 中国科学院计算技术研究所 | Message transmission frame and method based on high-speed network oriented to file system |
| CN101582058A (en) * | 2009-06-10 | 2009-11-18 | 华为技术有限公司 | Method and device for accessing registers |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030050990A1 (en) * | 2001-06-21 | 2003-03-13 | International Business Machines Corporation | PCI migration semantic storage I/O |
| US20060168091A1 (en) * | 2005-01-21 | 2006-07-27 | International Business Machines Corporation | RNIC-BASED OFFLOAD OF iSCSI DATA MOVEMENT FUNCTION BY INITIATOR |
| CN102571925B (en) * | 2011-12-19 | 2014-08-27 | 华中科技大学 | RDMA (Remote Direct Memory Access)-based data access method in fiber channel environment |
| CN103440202B (en) * | 2013-08-07 | 2016-12-28 | 华为技术有限公司 | A kind of communication means based on RDMA, system and communication equipment |
| CN103929415B (en) * | 2014-03-21 | 2018-03-06 | 华为技术有限公司 | The method, apparatus and network system of reading and writing data under RDMA |
-
2014
- 2014-03-21 CN CN201410108028.7A patent/CN103929415B/en active Active
- 2014-11-17 WO PCT/CN2014/091263 patent/WO2015139470A1/en active Application Filing
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1930563A (en) * | 2004-01-22 | 2007-03-14 | 高通股份有限公司 | A multiple address two channel bus structure |
| CN101459676A (en) * | 2008-12-31 | 2009-06-17 | 中国科学院计算技术研究所 | Message transmission frame and method based on high-speed network oriented to file system |
| CN101582058A (en) * | 2009-06-10 | 2009-11-18 | 华为技术有限公司 | Method and device for accessing registers |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2015139470A1 (en) | 2015-09-24 |
| CN103929415A (en) | 2014-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103929415B (en) | The method, apparatus and network system of reading and writing data under RDMA | |
| US9672143B2 (en) | Remote memory ring buffers in a cluster of data processing nodes | |
| CN109388595A (en) | High-bandwidth memory systems and logic dice | |
| CN104123262B (en) | The method and apparatus that the stream based on ID is enabled on PCIExpress | |
| US8239650B2 (en) | Wirelessly configurable memory device addressing | |
| US8606184B1 (en) | Coexistence message processing mechanism for wireless devices | |
| CN105335309B (en) | A kind of data transmission method and computer | |
| US8621030B2 (en) | Techniques to copy an operating system | |
| CN114356223B (en) | Memory access method and device, chip and electronic equipment | |
| TW201701194A (en) | Multi-frequency transponder | |
| CN109358974A (en) | A kind of method and relevant apparatus of interprocess communication | |
| CN110806997B (en) | System on chip and memory | |
| US20130054885A1 (en) | Multiport memory element and semiconductor device and system including the same | |
| CN110795382A (en) | Universal asynchronous receiving and transmitting transmitter based on FPGA and system on chip | |
| CN106959929A (en) | Memory and its method of work that a kind of multiport is accessed | |
| CN116860488A (en) | Process communication method and related device | |
| WO2019084789A1 (en) | Direct memory access controller, data reading method, and data writing method | |
| CN103399839B (en) | Media peripheral interface, electronic device and communication means | |
| CN103488597B (en) | First-in first-out buffer and data reading-writing method thereof | |
| JP2008547139A (en) | Method, apparatus and system for memory post-write buffer with unidirectional full-duplex interface | |
| CN106062814A (en) | Improved banked memory access efficiency by a graphics processor | |
| TW202230140A (en) | Method to manage memory and non-transitory computer-readable medium | |
| CN103647728B (en) | A kind of STM32 micro-chip and Linux system high-speed serial ports Anti-jamming Communication method | |
| CN207835492U (en) | A kind of Double buffer carrier wave demodulation system | |
| CN113422792A (en) | Data transmission method and device, electronic equipment and computer storage 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220214 Address after: 550025 Huawei cloud data center, jiaoxinggong Road, Qianzhong Avenue, Gui'an New District, Guiyang City, Guizhou Province Patentee after: Huawei Cloud Computing Technologies Co.,Ltd. Address before: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen Patentee before: HUAWEI TECHNOLOGIES Co.,Ltd. |
|
| TR01 | Transfer of patent right |