CN104410655B - The storage system and method for distributed network based on naming mechanism - Google Patents
The storage system and method for distributed network based on naming mechanism Download PDFInfo
- Publication number
- CN104410655B CN104410655B CN201410503667.3A CN201410503667A CN104410655B CN 104410655 B CN104410655 B CN 104410655B CN 201410503667 A CN201410503667 A CN 201410503667A CN 104410655 B CN104410655 B CN 104410655B
- Authority
- CN
- China
- Prior art keywords
- list item
- network
- name
- data
- local
- 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 44
- 230000007246 mechanism Effects 0.000 title claims abstract description 20
- 230000008569 process Effects 0.000 claims description 14
- 230000004044 response Effects 0.000 claims description 14
- 238000013461 design Methods 0.000 abstract description 10
- 238000013500 data storage Methods 0.000 abstract description 7
- 238000012545 processing Methods 0.000 abstract description 6
- 230000006855 networking Effects 0.000 abstract description 3
- 238000009825 accumulation Methods 0.000 abstract description 2
- 239000010410 layer Substances 0.000 description 12
- 238000004458 analytical method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- CIHZODPTTZIDPV-UHFFFAOYSA-N CS.CS Chemical compound CS.CS CIHZODPTTZIDPV-UHFFFAOYSA-N 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 238000005473 Guinier-Preston zone Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 239000012792 core layer Substances 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000003672 processing method 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]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention discloses a kind of storage system and method for the distributed network based on naming mechanism, network level node is identical with storage aspect node framework, local data processing, external request processing are carried out based on the node framework, and the processing that outside name data return, it is designed for uniformity in network level and accumulation layer in face of the data block of name, the distributed data-storage system of flattening.The system reference design of name data network (Named Data Networking NDN), and the memory mechanism of itself and file system is combined, form a kind of new distributed memory system framework.
Description
Technical field
The present invention relates to computer distributed system design field, and in particular to a kind of distributed network based on naming mechanism
The storage system and method for network.
Background technology
Based on the internet of TCP/IP network architecture, huge success is achieved so far, completes the whole world
Information interconnection and intercommunication, promoted the overall salary strategy in the world.Current TCP/IP architectures originate from state of the U.S. in 1969
The network for the entitled ARPAnet that Fang Bu Advanced Study Projects management board begins setting up, the purpose is to only by the several of the U.S.
Military and research is linked up with host computer.The purpose of design of internet most original that is to say two machine point-to-points of guarantee
Communication.But with internet demand and the development of application, current internet gradually communicates to large-scale content from point to points
Distribute to change.
Current internet also exposes subject matter：99% content is the block of information with name on network, this
A little information are not only produced by great institutions (big video website, news portal etc.), and terminal user also gradually produces substantial amounts of block of information,
Internet has become the carrier of block of information consumption and issue.Current TCP/IP frameworks are mainly the point-to-point of host-to-host
Pattern, it is unfavorable for the issue and diffusion of information.For ambulant support, IP network is in the support pair set with mobile device IP
That also does is undesirable.
In the big data epoch, how efficiently to transmit and handle large-scale data is the difficulty that major IT companies face
Topic.Current distributed system mainly uses the pattern of large-scale computer cluster to substitute traditional giant computer pattern.
The Google whole world possesses hundreds of thousands platform server altogether, per minute to receive several thousands of times search in the whole world altogether, and data are carried out one second
Within quick-searching and response.In a distributed system, the problem of a key it is data for being distributed in each main frame
Carry out unified management and access.How the metadata of mass data carries out efficient index and manages, the data between machine
How effectively to be transmitted.
In distributed mass data storage system, main flow is exactly the transmission for naming data block in network.Tradition
Distributed Cluster typically using distributed file system on existing generic-document system (such as ZFS file system)
(such as google GFS), network level typically use the network based on ICP/IP protocol.So in whole distributed field system
Many unnecessary redundancies can be produced in system：
First, it is that rope is carried out to the data block of storage medium storage in file system aspect, the main function of file system
Skirt is managed, and distributed file system, and secondary management is carried out soon to these name data in global level.In network level,
Data can be by stripping and slicing again in file system, and increases host address IP description data.In receiving terminal, the packet of chopping is also
Can be by from Combination nova into data block.The basic reason for causing these redundancies is because the design machine that network separates with file system
System, causes in different aspects, has different description mechanism for same data block.
The content of the invention
Network data and local data are stored in each network node in view of the above-mentioned problems, the present invention provides one kind
The distributed network storage system and data processing method of naming mechanism in one storage table.
To reach above-mentioned purpose, the distributed network storage system of the invention based on naming mechanism, each network node
Including at least local and network metadata table, described local comprises at least name prefix list item, network with network metadata table
Interface IP address list item and local address list item；
Wherein, the name prefix of the name prefix list item storage name data；
The network interface address of the network interface address list item storage name data；
The local address of the local address list item storage name data.
Further, described network node also includes content caching table and request waits table, wherein,
The data cached table stores the name data of the target node；
The request data package for waiting required list to store network node reception.
To reach above-mentioned purpose, the distributed network storage method of the invention based on naming mechanism, each network node
Including at least local and network metadata table, the local comprises at least name prefix list item with network metadata table, network connects
Port address list item and local address list item, wherein, the name prefix of the name prefix list item storage name data, the net
The network interface address of network interface IP address list item storage name data, the local of the local address list item storage name data
Address；
Described method comprises at least：
Receive the request bag for the name prefix for including name data；
The local request bag described with the response of network metadata table, wherein response method include：
The name prefix for the name data that name prefix in the request bag is stored with the name prefix list item enters
Row matching,
If there is occurrence, searched in the network interface address list item and/or the local address list item to should name
Claim the network interface address and/or local address of the name data of prefix；
Based on lookup result by should name prefix name data be back to the request bag send node；
If without occurrence, do not process.
Specifically, it is described in network interface address list item and/or the local described in the network interface address list item
The network interface address and/or local address of the name data are searched in address list item, specific steps include：
Judge whether the network interface address list item and local address list item have lookup priority,
If there is priority, searched based on priority in the high list item of priority；
If without priority, Network Search interface entry and local address list item simultaneously.
Specifically, it is described based on lookup result by should the name data of name prefix be back to the request bag
Node is sent, specific steps include：
If finding the network interface address of network naming data in network interface list item, by the request of reception
Bag is sent to described network interface address, and the network interface address returns to corresponding name data, exports the name data extremely
The request bag sends node；
If finding network naming data in local address list item, the name data are exported to the hair of the request bag
Egress；
If found in network interface list item and local address list item simultaneously, the priority based on result output is by net
Network interface IP address list item or the corresponding name data of local address list item output.
Further, described network node also includes content caching table and request waits table, and described method also includes：
The request bag request of reception is added into request to wait in table；
Content caching table respond request bag, wherein response method are：Search whether there is in described data cached table with
The name prefix that the name data name prefix that the request bag includes matches,
If in the presence of, the name data are sent to the node that sends of the request bag, and delete wait required list in phase
The wait request answered；
If being not present, search local with whether there is what is matched with name prefix in request bag in network metadata table
Name prefix,
If in the presence of, the local and network metadata table respond request bag, and removal request is waited in table and accordingly please
Seek bag；
If being not present, do not process.
The storage system and method for distributed network of the invention based on naming mechanism, by the way that the network for naming data is connect
Port address and local address are stored in same local and network metadata table so that in a network and the name that is locally stored
Data use same description mechanism, and name data can be searched on network level and storage aspect simultaneously by receiving request bag,
Name data are searched in storage aspect and need not move through extra data processing step, the search procedure of name data is simplified, carries
The high search efficiency of name data.
Brief description of the drawings
Fig. 1 is nodal analysis method schematic diagram of the present invention；
Fig. 2 is the global deployment schematic diagram of present system；
Fig. 3 is inventive network protocol stack structure.
Embodiment
With reference to Figure of description, the present invention will be further described.
The present invention proposes what a kind of data block for facing to name in network level and accumulation layer was designed for uniformity, flattening
Distributed data-storage system.The system reference the setting of name data network (Named Data Networking NDN)
Meter, and the memory mechanism of itself and file system is combined, a kind of new distributed memory system framework is formd, it is each
Individual network node is NDSS network nodes.
The present invention adds local and network metadata table, described local and network on the basis of data network is named
Metadata table comprises at least name prefix list item, network interface address list item and local address list item；
Wherein, the name prefix of the name prefix list item storage name data；
The network interface address of the network interface address list item storage name data；
The local address of the local address list item storage name data.
NDN projects rise in CCN (Content Cetntirc Networking content center networks) project.CCN projects
It is the pa Lip river atropic research center initiation of Xerox, its design philosophy is that the Van Jacobson in the U.S. are proposed.Main pin
Scalability and effective content distribution problem, the problem to current ICP/IP protocol have several years ago caused nerve of a covering
The research boom of (Overlay Network) and content distributing network (Content Delivery Network).But through excessive
Year application practice find that P2P and CDN still suffer from some shortcomings when solving the problems, such as content distribution, such as network/net neutrality with
The problem of large scale deployment.This new network architectures of NDN are intended to substitute IP layers using the name to data, turn into new
Network eel-like figure, i.e., layer network needs the standard that follows up and down.TCP/IP procotols are with NDN procotols to such as Fig. 1
It is shown.Wherein Content chuncks are to name data block.
(1) block data structure is named
Data block size changes according to the difference of storage medium.Such as a sector storage size is in general hard disk
4096B, then the size of data block is defaulted as 4096B in this distributed memory system.
Data block contents are divided into 4 parts, and Part I is the name of the data block.Data block naming method is using general
URI naming method, such as：ndn://tsinghua.edu.cn/video/intro.avi/s001/v001 or http://
portal.tsinghua.edu.cn/index.php/2013_11_10/v012.Name for data can be any, at that time
The general specification of URL names can be followed.
Part II is optional part, is digital signature, for Part I name, Part III signing messages and the 4th
The digital signature of partial data particular content.SHA-1 can be typically used, X.509 waits unsymmetrical key signature technology.
Part III is also optional part, is signing messages, including public key of signer etc..
Part IV is the concrete structure of the name data block.
Data block use can use xml codings TLV can also be used to encode, and here is a kind of example of coded format.
Wherein datablock is data block label, and name is data name label, and signature is digital signature label,
Signedinfo is signing messages label, and data is specific data label.
<datablock>
<name></name>
<signature></signature>
<signedinfo></signedinfo>
<data></data>
</datablock>
(2) data block request form is named
Name data block request is one section of message sequence, and request foremost is necessary for the name of data.Below can basis
The corresponding other fields of real needs increase.
(3) system node model
System node model reference NDN nodal analysis methods design, and local and network metadata table can be to local data
It is managed, nodal analysis method schematic diagram is as shown in Figure 2.
Wherein NDS Node are the nodal analysis method of the system, name data-storage system (Named Data Storage
System NDSS).The main task of nodal analysis method is, response upper level applications initiate for local or long-range
Request of data, response or the outside request for naming data of processing.If, should by inquiring about local metadata table (LFIB)
Data are present, then directly carry out inverse response according to request of data direction；If it does not exist, then request is forwarded according to LFIB.
LFIB is local and network metadata table (Local and Forwarding Information Base), for this
Ground is stored in network metadata table.The table mainly has three list items, name prefix (Prefix), interface table (Face list), sheet
Ground storage information (Local inode).Name prefix is similar to the IP prefixes of IP route table, is exactly one before several words
Point.Such as http://tsinghua.edu.cn/index/2013_11_10/v015, then one of which prefix is http://
tsinghua.edu.cn/.Interface table is used for storing, after the request for meeting the prefix stored in table, it should asked to which interface
Seek data.Such as the request of data of/parc.com/audios/ prefixes in figure, it should please to forward to 0 interface and 1 interface
Ask.Local inode are used for storing position of the name data block in local memory device.
Request waits table (Pending Request Table, PRT) and not responded to also for storing having arrived
Request.When have data block request reach node when, through inquiry LFIB locally without the data it is necessary to forward this data please
Ask.And the request of data can be also recorded in PRT.PRT data structure is the name of data, and the interface in request of data source
Title.
CS is content caching structure (Content Store), the name data block that can be returned with temporary cache NDSS.When
After node returns to the data block present in local or obtains the name data block that adjacent node is returned, the data block can be delayed
Exist in CS.CS data storage structure is name (name) and data block (block).The replacement policy of caching can be according to being
The real needs of system carry out individual cultivation.
The concept of interface is source and forwarding whereabouts of the request of data with naming data block.Remote interface is the network port,
Such as wired and wireless Ethernet interface.The local request that can directly handle local program.
The handling process locally asked is：
Local program sends name request of data to NDSS node procedures, and NDSS first looks for local CS cachings, if
Have and cache the name data block, then directly return to the data block.PRT tables are inquired about afterwards, the request if waiting identical data,
Then increase the request interface of PRT list items.No, then request of data is recorded in PRT tables.LFIB tables are searched afterwards.If
Locally there is the data block, is then directly returned to local program.According to the record of PRT tables, the request of wait is reversely sent
Data request interface (in addition to locally asking) is given, deletes the corresponding entry of PRT tables.Finally this is counted according to the cache policy configured
CS is put into according to block.
The handling process of external request is：
External request from certain interface to NDSS nodes after can search first it is local first look for local CS cachings, if
The name data block is cached, then directly returns to the data block.PRT tables are inquired about afterwards, the request if waiting identical data, then
Increase the request interface of PRT list items.No, then request of data is recorded in PRT tables.LFIB tables are searched afterwards.If at this
There is the data block on ground, according to the record of PRT tables, by the request of wait, is reversely sent to data request interface and deletes PRT tables
Corresponding entry.The data block is finally put into by CS according to the cache policy configured.If the not data block, according to LFIB tables
Face list network interfaces request is forwarded.
Outside name data return handling process be：
, can be according to PRT tables, by the interface of the reverse return recording of data block after external data block returns to NDSS nodes.It
The corresponding entry of PRT tables is deleted afterwards, and the data block is finally put into by CS according to the cache policy configured.
(4) system world model
Global system is intended to according to a set of distributed memory system prototype of foregoing memory node modelling.The system is by propping up
Hold agreement and property server or the virtual server composition of instruction that the system is followed.Each node supports identical to lead to
Letter association, using same nodal analysis method.But for an overall distribution formula group of planes, each responsibility is different from the division of labor, so each
Emphasis of the node in system design of dividing the work is different.If Fig. 3 is the global deployment schematic diagram of system.System is divided into three layers：
Core layer, predominantly NDSS core routers, Guinier-Preston zone, it is bis- layers of routers of NDSS, boundary layer, is NDSS clients or storage
Server.
Wherein NDSS Core Router are system core router, positioned at the core position of group of planes topology.It is mainly used to
Handle the forwarding and the return of name packet of substantial amounts of request of data.In system design, NDSS Core should have bigger
Forwarding packet ability.
NDSS L2 Router are two layers of router of system, undertake the request of data of boundary node and turning for name packet
Hair.Other its is mainly responsible for boundary node and cached, so, it is necessary to be its design comparison Large Copacity in specific implementation
CS。
NDSS Com are common border NDSS application servers, can be run above abundant based on NDSS system datas
The application of request protocol.NDSS Repo are NDSS data storage servers, and the persistent storage of Large Copacity can be designed for it.
It is attached between node using the procotol of following (5) description
(5) network connection
In order to ensure that the high-efficiency network between NDSS nodes transmits, grid upper-layer protocol does not use ICP/IP protocol,
But NDSS procotol will be directly run on link layer.Wherein so that link layer runs Ethernet protocol as an example
(Ethernet Frame).NDSS name data block (NDSS Chunks) is directly encapsulated in each ethernet frame.
Physical Layer are physical layer, that is, transmit the physical medium of network signal.
Present invention may apply to general commercial server cluster, and embodiment below is a kind of case study on implementation, other
Specific embodiment may be referred to the design in the content of the invention, be designed according to demand.
Server node is in this example, runs the general commercial server of (SuSE) Linux OS.Network level is directly run
Ethernet network agreement.The program for being responsible for NDSS processing data packets flows is operated on the server as kernel process.
Server storage device is the disk that sector-size is 4096B.Each sector directly stores a name data block.
The LFIB tables of NDSS processes directly carry out metadata management to the name data block on disk.The process can directly read hard disk fan
Data block in area.
Server via Ethernet card, by ethernet frame agreement packet will be named to be packaged and transmit.Due to one
As ethernet frame a length of 1500 byte of largest frames.In order to which an ethernet frame can encapsulate a complete name data block.
Need to be adjusted to Ethernet maximum frame length to be more than 4096 bytes.Many Ethernet switches and Ethernet card currently on the market
Support the Ethernet frame length more than 4096 bytes.
Embodiment 1
Distributed network storage system of the present embodiment based on naming mechanism, each network node comprise at least it is local with
Network metadata table, described local and network metadata table comprise at least name prefix list item, network interface address list item with
And local address list item；
Wherein, the name prefix of the name prefix list item storage name data；
The network interface address of the network interface address list item storage name data；
The local address of the local address list item storage name data.
Described network node also includes content caching table and request waits table, wherein,
The data cached table stores the name data of the target node；
The request data package for waiting required list to store network node reception.
The method of system respond request bag described in the present embodiment includes：
Receive the request bag for the name prefix for including name data；
The request bag request of reception is added into request to wait in table；
Content caching table respond request bag, wherein response method are：Search whether there is in described data cached table with
The name prefix that the name data name prefix that the request bag includes matches,
If in the presence of, the name data are sent to the node that sends of the request bag, and delete wait required list in phase
The wait request answered；
If being not present, search local with whether there is what is matched with name prefix in request bag in network metadata table
Name prefix,
If in the presence of, the local and network metadata table respond request bag, and removal request is waited in table and accordingly please
Seek bag；
If being not present, do not process.
The local and the method for network metadata table respond request bag comprise at least：
The name prefix for the name data that name prefix in the request bag is stored with the name prefix list item enters
Row matching,
If there is occurrence, searched in the network interface address list item and/or the local address list item to should name
Claim the network interface address and/or local address of the name data of prefix；
Based on lookup result by should name prefix name data be back to the request bag send node；
If without occurrence, do not process.
It is described in network interface address list item described in the network interface address list item and/or the local address list item
The middle network interface address and/or local address for searching the name data, specific steps include：
Judge whether the network interface address list item and local address list item have lookup priority,
If there is priority, searched based on priority in the high list item of priority；
If without priority, Network Search interface entry and local address list item simultaneously.
It is described based on lookup result by should name prefix name data be back to the request bag send section
Point, specific steps include：
If finding the network interface address of network naming data in network interface list item, by the request of reception
Bag is sent to described network interface address, and the network interface address returns to corresponding name data, exports the name data extremely
The request bag sends node；
If finding network naming data in local address list item, the name data are exported to the hair of the request bag
Egress；
If found in network interface list item and local address list item simultaneously, the priority based on result output is by net
Network interface IP address list item or the corresponding name data of local address list item output.
Embodiment 2
Distributed network storage method of the present embodiment based on naming mechanism, each network node comprise at least local
With network metadata table, the local and network metadata table comprise at least name prefix list item, network interface address list item with
And local address list item, wherein, the name prefix of the name prefix list item storage name data, the network interface address table
The network interface address of item storage name data, the local address of the local address list item storage name data；
Described method comprises at least：
Receive the request bag for the name prefix for including name data；
The local request bag described with the response of network metadata table, wherein response method include：
The name prefix for the name data that name prefix in the request bag is stored with the name prefix list item enters
Row matching,
If there is occurrence, searched in the network interface address list item and/or the local address list item to should name
Claim the network interface address and/or local address of the name data of prefix；
Based on lookup result by should name prefix name data be back to the request bag send node；
If without occurrence, do not process.
Further, it is described in network interface address list item described in the network interface address list item and/or described
The network interface address and/or local address of the name data are searched in way address list item, specific steps include：
Judge whether the network interface address list item and local address list item have lookup priority,
If there is priority, searched based on priority in the high list item of priority；
If without priority, Network Search interface entry and local address list item simultaneously.
Further, it is described based on lookup result by should the name data of name prefix be back to the request bag
Send node, specific steps include：
If finding the network interface address of network naming data in network interface list item, by the request of reception
Bag is sent to described network interface address, and the network interface address returns to corresponding name data, exports the name data extremely
The request bag sends node；
If finding network naming data in local address list item, the name data are exported to the hair of the request bag
Egress；
If found in network interface list item and local address list item simultaneously, the priority based on result output is by net
Network interface IP address list item or the corresponding name data of local address list item output.
Further, described network node also includes content caching table and request waits table, and described method also includes：
The request bag request of reception is added into request to wait in table；
Content caching table respond request bag, wherein response method are：Search whether there is in described data cached table with
The name prefix that the name data name prefix that the request bag includes matches,
If in the presence of, the name data are sent to the node that sends of the request bag, and delete wait required list in phase
The wait request answered；
If being not present, search local with whether there is what is matched with name prefix in request bag in network metadata table
Name prefix,
If in the presence of, the local and network metadata table respond request bag, and removal request is waited in table and accordingly please
Seek bag；
If being not present, do not process.
More than, only presently preferred embodiments of the present invention, but protection scope of the present invention is not limited thereto is any to be familiar with sheet
Those skilled in the art the invention discloses technical scope in, the change or replacement that can readily occur in should all be covered
Within protection scope of the present invention.Therefore, protection scope of the present invention should be defined by the protection domain that claim is defined.
Claims (6)
- A kind of 1. distributed network storage system based on naming mechanism, it is characterised in that：Each network node comprises at least It is local to comprise at least name prefix list item, network interface address with network metadata table with network metadata table, described local List item and local address list item；Wherein, the name prefix of the name prefix list item storage name data；The network interface address of the network interface address list item storage name data；The local address of the local address list item storage name data.
- 2. the distributed network storage system according to claim 1 based on naming mechanism, it is characterised in that：Described net Network node also includes content caching table and request waits table, wherein,The content caching table stores the name data of the target node；The request wait table stores the request data package of network node reception.
- A kind of 3. distributed network storage method based on naming mechanism, it is characterised in that：Each network node comprises at least It is local to comprise at least name prefix list item, network interface address table with network metadata table with network metadata table, the local Item and local address list item, wherein, the name prefix of the name prefix list item storage name data, the network interface The network interface address of location list item storage name data, the local address of the local address list item storage name data；Described method comprises at least：Receive the request bag for the name prefix for including name data；The local request bag described with the response of network metadata table, wherein response method include：By the name prefix progress of the name prefix in the request bag and the name data of name prefix list item storage Match somebody with somebody,If there is occurrence, searched in the network interface address list item and/or the local address list item to should be before title The network interface address and/or local address for the name data sewed；Based on lookup result by should name prefix name data be back to the request bag send node；If without occurrence, do not process.
- 4. the distributed network storage method according to claim 3 based on naming mechanism, it is characterised in that：The described network that the name data are searched in the network interface address list item and/or the local address list item connects Port address and/or local address, specific steps include：Judge whether the network interface address list item and local address list item have lookup priority,If there is priority, searched based on priority in the high list item of priority；If without priority, Network Search interface entry and local address list item simultaneously.
- 5. the distributed network storage method according to claim 3 based on naming mechanism, it is characterised in that：It is described based on lookup result by should the name data of name prefix be back to the node that sends of the request bag, tool Body step includes：If found in network interface list item name data network interface address, by the request bag of reception send to Described network interface address, the network interface address return to corresponding name data, export the name data to the request Bag sends node；If finding name data in local address list item, export the name data and send node to the request bag；If found in network interface list item and local address list item simultaneously, the priority based on result output is connect by network Port address list item or the corresponding name data of local address list item output.
- 6. the distributed network storage method according to claim 3 based on naming mechanism, it is characterised in that：Described net Network node also includes content caching table and request waits table, and described method also includes：The request bag request of reception is added into request to wait in table；Content caching table respond request bag, wherein response method are：Search whether there is in described data cached table with it is described The name prefix that the name data name prefix that request bag includes matches,If in the presence of, the name data are sent to the node that sends of the request bag, and delete wait it is corresponding in required list Wait request；If being not present, search local with whether there is the title to match with name prefix in request bag in network metadata table Prefix,If in the presence of, the local and network metadata table respond request bag, and removal request waits corresponding request bag in table；If being not present, do not process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410503667.3A CN104410655B (en) | 2014-09-26 | 2014-09-26 | The storage system and method for distributed network based on naming mechanism |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410503667.3A CN104410655B (en) | 2014-09-26 | 2014-09-26 | The storage system and method for distributed network based on naming mechanism |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104410655A CN104410655A (en) | 2015-03-11 |
| CN104410655B true CN104410655B (en) | 2018-02-09 |
Family
ID=52648253
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410503667.3A Active CN104410655B (en) | 2014-09-26 | 2014-09-26 | The storage system and method for distributed network based on naming mechanism |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104410655B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104780205B (en) * | 2015-03-24 | 2018-03-02 | 清华大学深圳研究生院 | The content requests and transmission method and system of content center network |
| CN104967677B (en) * | 2015-06-05 | 2019-07-19 | 无锡量子云数字新媒体科技有限公司 | A kind of document transmission method and device based on NDN cache optimization |
| CN106936909A (en) * | 2017-03-10 | 2017-07-07 | 北京工业大学 | A kind of method of Traffic information demonstration based on numerical nomenclature network with retrieving |
| CN108881071B (en) * | 2017-05-16 | 2020-02-14 | 华为技术有限公司 | Data packet transmission method and device storage medium under Ethernet content-centric networking mixing |
| CN109561355B (en) * | 2017-09-27 | 2020-07-17 | 中国科学院声学研究所 | System and method for CCN/NDN content registration, content location analysis and content routing |
| CN110138661A (en) * | 2019-04-02 | 2019-08-16 | 天津大学 | Name data network content storage pool neural network based |
| CN113114575A (en) * | 2021-03-26 | 2021-07-13 | 井芯微电子技术(天津)有限公司 | Network forwarding device and data forwarding method |
| CN114257536B (en) * | 2021-11-05 | 2023-09-01 | 浙江木链物联网科技有限公司 | Industrial data acquisition method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1914609A (en) * | 2003-12-12 | 2007-02-14 | 英国电讯有限公司 | Distributed computer system |
| CN101496005A (en) * | 2005-12-29 | 2009-07-29 | 亚马逊科技公司 | Distributed replica storage system with web services interface |
| CN103457999A (en) * | 2013-08-06 | 2013-12-18 | 北京大学深圳研究生院 | P2P document transmission method based on NDN network architecture |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8891541B2 (en) * | 2012-07-20 | 2014-11-18 | International Business Machines Corporation | Systems, methods and algorithms for named data network routing with path labeling |
| US8965845B2 (en) * | 2012-12-07 | 2015-02-24 | International Business Machines Corporation | Proactive data object replication in named data networks |
| US9049252B2 (en) * | 2012-12-17 | 2015-06-02 | International Business Machines Corporation | Efficient name management for named data networking in datacenter networks |
-
2014
- 2014-09-26 CN CN201410503667.3A patent/CN104410655B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1914609A (en) * | 2003-12-12 | 2007-02-14 | 英国电讯有限公司 | Distributed computer system |
| CN101496005A (en) * | 2005-12-29 | 2009-07-29 | 亚马逊科技公司 | Distributed replica storage system with web services interface |
| CN103457999A (en) * | 2013-08-06 | 2013-12-18 | 北京大学深圳研究生院 | P2P document transmission method based on NDN network architecture |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104410655A (en) | 2015-03-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104410655B (en) | The storage system and method for distributed network based on naming mechanism | |
| US10887279B2 (en) | Scalable cloud hosted metadata service | |
| CN102387061B (en) | Method, device and system for accessing VPC (virtual private cloud) to VPN (virtual private network) | |
| CN105868231A (en) | Cache data updating method and device | |
| CN101465889A (en) | Network address translation equipment and request method of response address analysis protocol | |
| CN106156359A (en) | A kind of data synchronization updating method under cloud computing platform | |
| US20200159775A1 (en) | Network-wide, location-independent object identifiers for high-performance distributed graph databases | |
| US10230802B2 (en) | Providing stateless network services | |
| CN102413041B (en) | Method, device and system for moving security policy | |
| CN103366117B (en) | A kind of viral restorative procedure of infection type and system | |
| WO2016146011A1 (en) | Method, system and management system for constructing virtual non-volatile storage medium | |
| CN102438020A (en) | Method and equipment for distributing contents in content distribution network, and network system | |
| JP2015509232A (en) | Method and apparatus for messaging in the cloud | |
| US9311314B2 (en) | System and method for migrating data from a source file system to a destination file system with use of attribute manipulation | |
| CN108287905A (en) | A kind of extraction of network flow feature and storage method | |
| CN109583221A (en) | Dropbox system based on cloudy server architecture | |
| CN108270878A (en) | Method, the VTEP equipment of ARP messages are sent in VXLAN | |
| CN104639592A (en) | Cloud storage gateway and method and system for information synchronization among cloud storage gateways | |
| JP2017500679A (en) | Media resource feedback method, apparatus, program, and recording medium | |
| CN103559279A (en) | Cloud storage acceleration method and device | |
| CN104468822A (en) | Media platform architecture for cloud calculation | |
| Schales et al. | FCCE: highly scalable distributed feature collection and correlation engine for low latency big data analytics | |
| Song et al. | Exploring Reliable Decentralized Networks with Smart Collaborative Theory | |
| CN102185935B (en) | Resource-oriented hierarchical name resolution system | |
| WO2024103800A1 (en) | Method and apparatus for live migration of virtual machine, and device and non-volatile readable 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 |