CN1674538A - Network storing system based on local network and its method for reading and writing data - Google Patents
Network storing system based on local network and its method for reading and writing data Download PDFInfo
- Publication number
- CN1674538A CN1674538A CN 200410008530 CN200410008530A CN1674538A CN 1674538 A CN1674538 A CN 1674538A CN 200410008530 CN200410008530 CN 200410008530 CN 200410008530 A CN200410008530 A CN 200410008530A CN 1674538 A CN1674538 A CN 1674538A
- Authority
- CN
- China
- Prior art keywords
- layer
- network
- data
- server
- scsi
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to a network storage system based on local area network and its data readring and writing method. Said system is a machine group formed from two types of servers which are mutually connected by means of local area network; in which one type of server is network magnetic disk server, every network magnetic disk server is connected with high-capacity magnetic disk equipment for storing data, and another type of server is node computer, every node computer not only can be connected with network magnetic disk server by means of internal local area network, but also can be connected with external customer computer outside of system by means of external network. Besides, said invention also provides the concrete action of above-mentioned two types of servers and application of said network storage system.
Description
Technical field
The present invention relates to a kind of storage system and method for reading and writing data thereof, belong to the computer network system technical field of structures based on local area network (LAN).
Background technology
Along with the exponential growth of data volume, storage becomes very important problem.The data storage that traditional server adopts at capacity, share, there are many shortcomings in aspect such as autgmentability, can't satisfy the demand of current storage system.The network store system of Chu Xianing makes the storage problem be resolved in recent years.
At present, main network store system has two kinds of structures: a kind of structure is storage area network (SAN, Storage Area Network), it is to use special fiber channel network (FC, Fibre Channel) server and optical fiber memory device are coupled together, server on the FC network can be visited the optical fiber memory device as the visit local disk, and block-based interface is provided.Another structure is called network attached storage (NAS, Network Attached Storage), it is to utilize an existing local area network (LAN) and a special server on network to dispose jumbo disk unit, and the establishment file system is formed on these memory devices.This server can pass through NFS (NFS, Network File System), the internet public documents (CIFS of system, Internet Common File System) etc. the file sharing protocol of industrial standard is exported to other machines on the network with this document system, thereby realizes the data sharing of file one-level.Because SAN uses dedicated network, so excellent performance, the SAN structure also has fabulous autgmentability, buys the new optical fiber memory device existing system of just can upgrading.And NAS has also obtained to accept widely and use because dispose advantage such as simple, cheap, that management is easy.
But these two kinds of network store system structures of SAN and NAS also have new problem: since traditional SAN exist cost an arm and a leg, shortcomings such as complex management, interoperability difference, make its range of application be restricted, be confined to the solid financial strength large enterprises of minority.But numerous medium-sized and small enterprises exist the application demand to the disk storage of piece level equally, for example database etc.Since medium-sized and small enterprises widespread deployment local area network (LAN), and the technical staff has the use experience of safeguarding of abundant local area network (LAN) aspect, if therefore traditional FC network can be substituted with local area network (LAN), but still provide piece level memory interface, cost of investment will reduce greatly so, is certain to be subjected to numerous users' welcome.Therefore how in local area network (LAN), to realize special fiber channel network storage organization and using method thereof are subjected to the insider with regard to becoming new research topic concern.
Summary of the invention
The purpose of this invention is to provide a kind of network store system based on local area network (LAN), this system is based on local area network (LAN), can provide the interface of disk block one-level to carry out the reading and writing data visit with low price, simultaneously, this network store system has extraordinary extended capability, can reduce user's initial outlay, when service needed, expand again; And do not needing substantially to increase under the prerequisite of investment, its serviceability also has a certain upgrade than NAS.
Another object of the present invention provides a kind of data read-write method of the network store system based on local area network (LAN).
The object of the present invention is achieved like this: a kind of network store system based on local area network (LAN), and this system is the group of planes that the server that connects each other by local area network (LAN) is formed; It is characterized in that: described server is divided into two classes, and wherein a class server is called as the network disk server, and each network disk server is connected with the large capacity disc equipment that is used to store data; Another kind of server is called as the node machine, each node machine both connected the network disk server by internal lan, pass through the client computer of external network connected system outside again, each node machine is the independent file server that file service is provided to the external client, and node machine and network disk server are formed a group of planes integral body by local area network (LAN), finish file service request jointly from group of planes external client; Internal communication between this intrasystem network disk server and the node machine adopts towards the host-host protocol of storage.
The transmission rate of described local area network (LAN) should be not less than 100 MBPSs.
Described node machine is the host-host protocol of a kind of operational network file system, cluster file system, volume management software and this internal system communication, and promptly towards the computer of the host-host protocol of storing, this node machine is born the processing and the transmission of storage data.
Described network disk server is a kind of computer that has jumbo disk unit, and described disk unit is modeled into the local disk of each node machine on the group of planes by software.
The cluster nodes machine and/or the network disk server of described system all can be expanded according to business demand, with the load capacity that improves system and/or memory capacity, read or write speed.
Described host-host protocol towards storage is a kind of connection-oriented transport layer protocol, and this agreement does not have network layer, and route can not be provided between network; Its bottom is a link layer, directly utilizes link layer to realize two data transfer between devices of internal system, reducing the data processing of agreement, and provides reliable data transport service for upper layer application.
Described towards the BSD socket API layer (API of the host-host protocol of storing by standard, Application Programming Interface) and the protocol transmission layer form, this socket API function is called on the upper strata, can realize reliably transfer of data end to end, and simplify the graft application on upper strata; Promptly in application layer, system uses internet small computer system interface (iSCSI, Internet SCSI) protocol transmission small computer system interface (SCSI, Small Computer System Interface) order, state and data, transmit leg calls the service that this this protocol transmission layer of socket API function access provides when transfer of data, the iSCSI packet inerrably is sent to the recipient fast.
Described host-host protocol towards storage adopts with the group to be window and to adopt the mode of different rates to carry out flow control between group and group, with application and the raising data throughput that adapts to storage.
Another object of the present invention is achieved in that a kind of method for reading and writing data of the network store system based on local area network (LAN), it is characterized in that: comprise the following steps:
(1) client computer of system outside proposes the request that reads and writes data by external network to system;
(2) the node machine of system is handled client requests, and this request is sent to corresponding network storage server;
(3) network storage server is handled this request of reading and writing data, and result is sent back to corresponding node machine;
(4) the node machine is handled object information, and gives client computer with final result by external network.
Described step (2) further comprises the following steps:
(21) the some nodes in the group of planes are received reading and writing data when request of external client, and the NFS layer of this cluster nodes is accepted user's request, change it into to cluster file system request, give the cluster file system layer and handle;
(22) the cluster file system layer is handled this upper-layer user's request, searches the metadata information of user file, at the device number and the logical block number (LBN) that obtain storing this document, and behind the number of the data block that will read and write, gives the volume management software layer of lower floor with it;
(23) the volume management software layer is converted to physical device number and physical block number according to the configuration information of being preserved with device number and logical block number (LBN), and gives SCSI Drive Layer in this node machine with these information;
(24) the SCSI Drive Layer is converted to corresponding SCSI read write command with this user's request, and finds corresponding handling procedure, just iSCSI application layer according to physical device number;
(25) the iSCSI application layer encapsulates this scsi command, and transport layer is given in the order of calling after socket API function will encapsulate;
(26) transport layer adds the iSCSI packet to repack after packet header and other relevant informations and is transmits data packets, gives link layer with it again;
(27) link layer adds the frame head of oneself before transmits data packets, is the signal of telecommunication by physical layer conversion, sends to corresponding network storage server.
Described step (3) further comprises the following steps:
(31) physical layer of network storage server receives the transmits data packets signal, changes it into Frame, gives link layer;
(32) link layer removes frame head, after simply checking, gives transport layer with transmits data packets;
(33) transport layer is removed transmission packet header, takes out the iSCSI packet, gives iSCSI application layer by socket API layer again;
(34) the iSCSI application layer is taken out scsi command from the iSCSI packet, passes to the SCSI Drive Layer then, by the SCSI Drive Layer scsi command is sent to the disk unit driver;
(35) the disk unit driver is according to the physical block number that is comprised in the scsi command, and the data block that read-write specifies number is encapsulated in it in scsi data bag then, returns to the SCSI Drive Layer;
(36) the SCSI Drive Layer passes to the iSCSI application layer with the scsi data bag that returns, and by the iSCSI application layer scsi data bag is packaged into the iSCSI packet again, and calls socket API layer and give transport layer and transmit;
(37) transport layer adds that with the iSCSI packet transmission head and check information become transmits data packets, give link layer;
(38) link layer is encapsulated in transmits data packets in the Frame again, gives physical layer;
(39) physical layer is converted to the signal of telecommunication with Frame, sends to node machine physical layer corresponding in the group of planes.
Described step (4) further comprises the following steps:
(41) node machine physical layer receives the signal of telecommunication, is converted into Frame, gives link layer process;
(42) link layer removes frame head and check information, takes out transmits data packets, upwards gives transport layer;
(43) transport layer is taken out the iSCSI packet from transmits data packets, returns to the iSCSI application layer via socket API layer;
(44) the iSCSI application layer is taken out the scsi data bag from the iSCSI packet, gives SCSI Drive Layer with it;
(45) the SCSI Drive Layer is taken out the object information that reads and writes data in the scsi data bag, gives volume management software;
(46) the volume management software object information that will read and write data returns to cluster file system;
(47) the cluster file system object information that will read and write data returns to NFS;
(48) NFS returns to client computer by the external network object information that will read and write data.
The present invention is a kind of network store system that is different from traditional SAN and NAS structure, this system utilizes extensive use and cheap local area network (LAN) substituted for optical fibers passage, make the cost of investment mounting blocks level storage system reduce greatly, and have potential user quite widely.Another characteristic of system of the present invention is to propose a kind of agreement of dedicated transmissions efficiently-towards the host-host protocol of storage, the protocol processes time when service that this agreement directly uses link layer of local area network to provide reduces transfer of data, and the flow control mechanism that adopts window and speed dual mode to combine improves efficiency of transmission, makes the performance of this network store system be highly improved.Moreover system of the present invention can increase the load capacity that cluster nodes is come expanding system easily, and increases memory capacity and performance that the network disk server comes expanding system.This system realizes the single cluster file system map that the overall situation is consistent, and the client can have access to intrasystem total data by any one server node, thereby realizes load balancing.In a word, system configuration of the present invention is simple, and is with low cost, and reading and writing data is convenient quick, and reliable operation has good application prospects.
Description of drawings
Fig. 1 is that the structure that the present invention is based on the network store system of local area network (LAN) is formed schematic diagram.
Fig. 2 is the host-host protocol structural representation towards storage that internal system communication of the present invention is used.
Fig. 3 is the software layer aggregated(particle) structure schematic diagram in node machine of the present invention and the network storage server.
To be the present invention leave data on the n platform network storage server schematic diagram to Fig. 4 with distributing.
Embodiment
At present, the communication of local area network (LAN) storage system is mainly based on ICP/IP protocol, and the processing expenditure of ICP/IP protocol is quite huge, the so just serious raising that hinders systematic function.For this reason, the present invention proposes a kind of novel network store system based on local area network (LAN).And, the communication protocol that the internal communication between this intrasystem network disk server 1 and the node machine 2 adopts the present invention to propose towards storage.
Referring to Fig. 1, the network store system that the present invention is based on local area network (LAN) is the group of planes that the two class servers 1,2 that connect each other by local area network (LAN) 3 are formed, wherein a class server is called as network disk server 1, and each network disk server 1 is connected with the large capacity disc equipment that is used to store data; Another kind of server is the node machine 2 that file service is provided for the external client, each node machine 2 connects each network disk server 1 by the high-speed local area network 3 that internal transmission speed is not less than 100 MBPSs, external client 5 is connected to node machine 2 in the group of planes by external network 4, thereby can have access to the file system on the whole group of planes.Because each node machine 2 all is the file server that file service independently is provided to the external client, and by local area network (LAN) 3 each node machine 2 and network disk server 1 are formed a group of planes, group of planes integral body provides the file read-write access services to the external client jointly.
Node machine 2 of the present invention is host-host protocols of a kind of operational network file system, cluster file system, volume management software and this internal system communication, promptly towards the computer (referring to Fig. 3) of the host-host protocol of storing, this node machine is born the processing and the transmission of storage data.Network disk server 1 is a kind of computer that has jumbo disk unit, and these disk units are modeled into the local disk of each node machine 2 on the group of planes by software.And cluster nodes machine 2, the network disk server 1 of this system all can be expanded easily according to business demand, with performances such as the load capacity that improves system, memory capacity, read or write speeds.
The present invention is a kind of connection-oriented transport layer protocol towards the communication protocol (SOP, Storage Oriented Protocol) of storage, is used to provide reliable transfer of data.Relation between this SOP agreement and other protocol layers as shown in Figure 2.The transport layer of this agreement is played the part of very important role in the reading and writing data access process, the overall performance that its performance can decision systems.Because applied environment of the present invention is a local area network (LAN), the visit of data is in same network system, just in same subnet, so this agreement does not have network layer, can not provide route between network; The service that its transport layer directly utilizes link layer to provide realizes transfer of data between two machines.This agreement is the BSD socket API layer (API by standard, Application Programming Interface) and the protocol transmission layer form, this socket API function is called on the upper strata, can realize reliably transfer of data end to end, and simplify the graft application on upper strata; Just in application layer, system uses internet small computer system interface (iSCSI, Internet SCSI) protocol transmission small computer system interface (SCSI, Small Computer System Interface) order, state and data.When transfer of data, transmit leg calls the service that this this protocol transmission layer of socket API function access provides, and the iSCSI packet inerrably is sent to the recipient fast.Because this agreement has been reduced protocol hierarchy, make the processing expenditure of packet reduce greatly.
The present invention adopts with the group towards the host-host protocol of storage and is window-unit and combines at the dual mode that adopts different rates control between group and the group and to carry out flow control, adapts to the application and the raising data throughput of storage.The data envelope that wherein earlier all will be sent to the other side is dressed up many small data packets, again these small data packets is divided into several groups, in each group a plurality of (N) small data packets is arranged.Sending data is to be unit with the group, and promptly same group of interior small data packets is to send continuously; Between group and the group time out T is arranged, the recipient also is to be unit with the group for the affirmation of packet: when only receiving all packets in this group, ability is returned a confirmation of receipt to this group to transmit leg.Transmit leg just should be organized the shared buffering area of packet and discharge after receiving some group affirmation.Transmit leg can send W group data (being that the group window size is W) at most before the affirmation of receiving one group of packet.If sent W group data, still do not receive affirmation, transmit leg will suspend transmission, till receiving affirmation.Every group small data packets number N, group are that both sides consult to determine when being connected at first foundation with time interval T and group window size W between the group, but these parameters also can be changed during connecting, both sides can hold consultation again according to the network condition and the transmission situation of reality, and dynamically adjust, thereby reduce the probability that network congestion takes place, improve efficiency of transmission.
Referring to Fig. 3, introduce software layer aggregated(particle) structure and data access process on cluster nodes machine 2 of the present invention and the network disk server 1.From the upper strata, node machine 2 comprises NFS layer 6, cluster file system layer 7, volume management software layer 8, SCSI Drive Layer 9, iSCSI layer 10, SOP socket API11, SOP transport layer 12, link layer 13 successively, and physical layer 14.And the software and hardware hierarchical structure on network disk server 1 is: make progress from low layer, comprise physical layer 15, link layer 16, SOP transport layer 17, SOP socket API 19, iSCSI layer 19, SCSI Drive Layer 20 successively, and disk unit drives 21.
The process that external client 5 visits are stored in certain file on network disk server 1 disk in the system is such:
(1) client computer 5 of system outside proposes access request by external network 4 with the some node machines 2 in system's group of planes;
When (2) the some node machines 2 in the group of planes were received the access request of external client 5, the NFS layer 6 of this node was accepted user request, changed it into to cluster file system request, gave cluster file system layer 7 and handled;
(3) request on 7 pairs of upper strata of cluster file system layer is handled, and searches the metadata information of user file, finally obtains storing the device number and the logical block number (LBN) of this document, and the number of the data block that will read.Give the volume management software layer 8 of lower floor with it;
(4) volume management software layer 8 is converted to physical device number and physical block number according to the configuration information of being preserved with device number and logical block number (LBN), and gives SCSI Drive Layer 9 in this node machine with these information;
(5) SCSI Drive Layer 9 is converted to corresponding SCSI read command with this user's request, and finds corresponding handling procedure according to physical device number, and just the iSCSI application layer 10;
(6) iSCSI application layer 10 these scsi commands of encapsulation, and transport layer 12 is given in the order of calling after socket API function 11 will encapsulate;
(7) transport layer 12 adds the iSCSI packet to repack after packet header and other relevant informations and is transmits data packets, it is given link layer 13 again;
(8) link layer 13 adds the frame head of oneself before transmits data packets, is converted to the signal of telecommunication by physical layer 14, sends to corresponding network storage server 1.
(9) physical layer 15 of network storage server 1 receives the transmits data packets signal, changes it into Frame, gives link layer 16;
(10) link layer 16 removes frame head, after simply checking, gives transport layer 17 with transmits data packets;
(11) transport layer 17 is removed transmission packet header, takes out the iSCSI packet, gives iSCSI application layer 19 by socket API layer 18 again and handles;
(12) iSCSI application layer 19 is taken out scsi command from the iSCSI packet, passes to SCSI Drive Layer 20 then, by SCSI Drive Layer 20 scsi command is sent to disk unit driver 21;
(13) disk unit driver 21 is according to the physical block number that is comprised in the scsi command, and the data block that read-write specifies number is encapsulated in it in scsi data bag then, returns to SCSI Drive Layer 20;
(14) SCSI Drive Layer 20 passes to iSCSI application layer 19 with the scsi data bag that returns, and by iSCSI application layer 19 the scsi data bag is packaged into the iSCSI packet again, and calls socket API layer 18 and give transport layer 17 and transmit;
(15) transport layer 17 adds that with the iSCSI packet transmission head and check information become transmits data packets, give link layer 16;
(16) link layer 16 is encapsulated in transmits data packets in the Frame again, gives physical layer 15;
(17) physical layer 15 is converted to the signal of telecommunication with Frame, sends to node machine 2 physical layers 14 corresponding in the group of planes;
(18) physical layer 14 of node machine 2 receives the signal of telecommunication, is converted into Frame, gives link layer 13 and handles;
(19) link layer 13 removes frame head and check information, takes out transmits data packets, upwards gives transport layer 12;
(20) transport layer 12 is taken out the iSCSI packet from transmits data packets, returns to iSCSI application layer 10 via socket API layer 11;
(21) iSCSI application layer 10 is taken out the scsi data bag from the iSCSI packet, gives SCSI Drive Layer 9 with it;
(22) data message in the SCSI Drive Layer 9 taking-up scsi data bags is given volume management software 8;
(23) volume management software 8 returns to cluster file system 7 with data;
(24) cluster file system 7 returns to NFS 6 with data;
(25) NFS 6 returns to client computer 5 by external network 4 with data message.
Client computer 5 writes file in system of the present invention process also is similarly, does not repeat them here.
The present invention is by specifying the address of different network storage servers 1 in SOP transport layer 12, the node machine 2 in the group of planes is the disk on all network storage server 1 in can access system.Because the status of all cluster nodes machines 2 is equal to, so any one node machine 2 can be visited disk on any one network storage server in the system, and this visit is carried out simultaneously.
Referring to Fig. 4, cluster nodes machine 2 can leave data on n the network storage server 1 in the itemize mode, file is that unit is deposited with the data block, the 1st is placed on the 1st network storage server, the 2nd is placed on the 2nd network storage server, the n piece is placed on n the network storage server, and the n+1 piece is placed on again on the 1st server, and so circulation is deposited.Like this, when access file, can from this n network storage server, read continuous n blocks of data simultaneously at most, make the data bandwidth maximum can reach n times of single network storage server word length.This mode makes the performance of system of the present invention improve greatly.On the other hand, because all cluster nodes machines 2 can have access to intrasystem any one network storage server 1, and owing to there is the consistent single cluster file system of the overall situation, the external client visits any one cluster nodes machine 2 can obtain its needed data.Like this, external client's visit can be distributed to equably on all cluster nodes machines 2, thereby make the visit load on all node machines 2 reach balanced, it is very heavy some cluster nodes machine visit loads can not occur, and the very light phenomenon of the load of other node machines.By increasing the cluster nodes machine, just can be so that the load capacity of system is improved.Therefore the expansion of systematic function of the present invention comprises two aspects: the one, improve data throughput by increasing network storage server 1, and the 2nd, improve and be used to ask disposal ability by increasing cluster nodes machine 2.In addition, the expansion for capacity also can realize by increase network storage server in system.Therefore the expansion of system of the present invention is very simple and effective.
Claims (12)
1, a kind of network store system based on local area network (LAN), this system is the group of planes that the server that connects each other by local area network (LAN) is formed; It is characterized in that: described server is divided into two classes, and wherein a class server is called as the network disk server, and each network disk server is connected with the large capacity disc equipment that is used to store data; Another kind of server is called as the node machine, each node machine both connected the network disk server by internal lan, pass through the client computer of external network connected system outside again, each node machine is the independent file server that file service is provided to the external client, and node machine and network disk server are formed a group of planes integral body by local area network (LAN), finish file service request jointly from group of planes external client; Internal communication between this intrasystem network disk server and the node machine adopts towards the host-host protocol of storage.
2, network store system according to claim 1 is characterized in that: the transmission rate of described local area network (LAN) should be not less than 100 MBPSs.
3, network store system according to claim 1, it is characterized in that: described node machine is the host-host protocol of a kind of operational network file system, cluster file system, volume management software and this internal system communication, promptly towards the computer of the host-host protocol of storing, this node machine is born the processing and the transmission of storage data.
4, network store system according to claim 1 is characterized in that: described network disk server is a kind of computer that has jumbo disk unit, and described disk unit is modeled into the local disk of each node machine on the group of planes by software.
5, according to claim 3 or 4 described network store systems, it is characterized in that: the cluster nodes machine and/or the network disk server of described system all can be expanded according to business demand, with the load capacity that improves system and/or memory capacity, read or write speed.
6, network store system according to claim 1 is characterized in that: described host-host protocol towards storage is a kind of connection-oriented transport layer protocol, and this agreement does not have network layer, and route can not be provided between network; Its bottom is a link layer, directly utilizes link layer to realize two data transfer between devices of internal system, reducing the data processing of agreement, and provides reliable data transport service for upper layer application.
7, network store system according to claim 6, it is characterized in that: described host-host protocol towards storage is made up of the BSD socket API layer API and the protocol transmission layer of standard, this socket API function is called on the upper strata, can realize reliably transfer of data end to end, and simplify the graft application on upper strata; Promptly in application layer, system uses internet small computer system interface iSCSI protocol transmission small computer system interface scsi command, state and data, transmit leg calls the service that this this protocol transmission layer of socket API function access provides when transfer of data, the iSCSI packet inerrably is sent to the recipient fast.
8, network store system according to claim 6, it is characterized in that: described host-host protocol towards storage adopts with the group to be window and to adopt the mode of different rates to carry out flow control between group and group, with application and the raising data throughput that adapts to storage.
9, a kind of method for reading and writing data of the network store system based on local area network (LAN) is characterized in that: comprise the following steps:
(1) client computer of system outside proposes the request that reads and writes data by external network to system;
(2) the node machine of system is handled client requests, and this request is sent to corresponding network storage server;
(3) network storage server is handled this request of reading and writing data, and result is sent back to corresponding node machine;
(4) the node machine is handled object information, and gives client computer with final result by external network.
10, method according to claim 9 is characterized in that: described step (2) further comprises the following steps:
(21) the some nodes in the group of planes are received reading and writing data when request of external client, and the NFS layer of this cluster nodes is accepted user's request, change it into to cluster file system request, give the cluster file system layer and handle;
(22) the cluster file system layer is handled this upper-layer user's request, searches the metadata information of user file, at the device number and the logical block number (LBN) that obtain storing this document, and behind the number of the data block that will read and write, gives the volume management software layer of lower floor with it;
(23) the volume management software layer is converted to physical device number and physical block number according to the configuration information of being preserved with device number and logical block number (LBN), and gives SCSI Drive Layer in this node machine with these information;
(24) the SCSI Drive Layer is converted to corresponding SCSI read write command with this user's request, and finds corresponding handling procedure, just iSCSI application layer according to physical device number;
(25) the iSCSI application layer encapsulates this scsi command, and transport layer is given in the order of calling after socket API function will encapsulate;
(26) transport layer adds the iSCSI packet to repack after packet header and other relevant informations and is transmits data packets, gives link layer with it again;
(27) link layer adds the frame head of oneself before transmits data packets, is the signal of telecommunication by physical layer conversion, sends to corresponding network storage server.
11, method according to claim 9 is characterized in that: described step (3) further comprises the following steps:
(31) physical layer of network storage server receives the transmits data packets signal, changes it into Frame, gives link layer;
(32) link layer removes frame head, after simply checking, gives transport layer with transmits data packets;
(33) transport layer is removed transmission packet header, takes out the iSCSI packet, gives iSCSI application layer by socket API layer again;
(34) the iSCSI application layer is taken out scsi command from the iSCSI packet, passes to the SCSI Drive Layer then, by the SCSI Drive Layer scsi command is sent to the disk unit driver;
(35) the disk unit driver is according to the physical block number that is comprised in the scsi command, and the data block that read-write specifies number is encapsulated in it in scsi data bag then, returns to the SCSI Drive Layer;
(36) the SCSI Drive Layer passes to the iSCSI application layer with the scsi data bag that returns, and by the iSCSI application layer scsi data bag is packaged into the iSCSI packet again, and calls socket API layer and give transport layer and transmit;
(37) transport layer adds that with the iSCSI packet transmission head and check information become transmits data packets, give link layer;
(38) link layer is encapsulated in transmits data packets in the Frame again, gives physical layer;
(39) physical layer is converted to the signal of telecommunication with Frame, sends to node machine physical layer corresponding in the group of planes.
12, method according to claim 9 is characterized in that: described step (4) further comprises the following steps:
(41) node machine physical layer receives the signal of telecommunication, is converted into Frame, gives link layer process;
(42) link layer removes frame head and check information, takes out transmits data packets, upwards gives transport layer;
(43) transport layer is taken out the iSCSI packet from transmits data packets, returns to the iSCSI application layer via socket API layer;
(44) the iSCSI application layer is taken out the scsi data bag from the iSCSI packet, gives SCSI Drive Layer with it;
(45) the SCSI Drive Layer is taken out the object information that reads and writes data in the scsi data bag, gives volume management software;
(46) the volume management software object information that will read and write data returns to cluster file system;
(47) the cluster file system object information that will read and write data returns to NFS;
(48) NFS returns to client computer by the external network object information that will read and write data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410008530 CN1674538A (en) | 2004-03-25 | 2004-03-25 | Network storing system based on local network and its method for reading and writing data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410008530 CN1674538A (en) | 2004-03-25 | 2004-03-25 | Network storing system based on local network and its method for reading and writing data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1674538A true CN1674538A (en) | 2005-09-28 |
Family
ID=35046815
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410008530 Pending CN1674538A (en) | 2004-03-25 | 2004-03-25 | Network storing system based on local network and its method for reading and writing data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1674538A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100442780C (en) * | 2006-01-27 | 2008-12-10 | 杭州华三通信技术有限公司 | iSCSI Storage Network, Disk Array and Its Capacity Expansion Method |
| CN102263818A (en) * | 2011-07-07 | 2011-11-30 | 北京飞杰信息技术有限公司 | Method for storing and reading file data, and apparatus thereof |
| US8132033B2 (en) | 2008-07-07 | 2012-03-06 | Hitachi, Ltd. | Storage system |
| CN101763051B (en) * | 2008-12-24 | 2012-07-25 | 台达电子工业股份有限公司 | Monitor Machine Method |
| CN102693151A (en) * | 2011-03-23 | 2012-09-26 | 鸿富锦精密工业（深圳）有限公司 | Computer and data downloading method thereof |
| CN102929791A (en) * | 2012-10-25 | 2013-02-13 | 浪潮电子信息产业股份有限公司 | Cluster storage system expansion method based on identical expansion module |
| CN101356783B (en) * | 2006-01-12 | 2014-06-04 | 博通以色列研发公司 | Method and system for protocol offload and direct I/O with I/O sharing in a virtualized network environment |
| CN104796774A (en) * | 2015-03-23 | 2015-07-22 | 四川长虹电器股份有限公司 | Across-device application integrating method and system |
| CN105471927A (en) * | 2014-05-30 | 2016-04-06 | 华为技术有限公司 | Method for processing service routing message, and service node |
| CN108551477A (en) * | 2018-03-28 | 2018-09-18 | 深圳市网心科技有限公司 | Data transmission channel establishment system, network storage device, server and method |
| CN117376289A (en) * | 2023-10-11 | 2024-01-09 | 哈尔滨工业大学 | Network disk data scheduling method for road monitoring data usage application |
| CN119127798A (en) * | 2024-11-13 | 2024-12-13 | 杭州浩联智能科技有限公司 | A file storage method, device, equipment and storage medium |
-
2004
- 2004-03-25 CN CN 200410008530 patent/CN1674538A/en active Pending
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101356783B (en) * | 2006-01-12 | 2014-06-04 | 博通以色列研发公司 | Method and system for protocol offload and direct I/O with I/O sharing in a virtualized network environment |
| CN100442780C (en) * | 2006-01-27 | 2008-12-10 | 杭州华三通信技术有限公司 | iSCSI Storage Network, Disk Array and Its Capacity Expansion Method |
| US8132033B2 (en) | 2008-07-07 | 2012-03-06 | Hitachi, Ltd. | Storage system |
| US8375235B2 (en) | 2008-07-07 | 2013-02-12 | Hitachi, Ltd. | Storage system |
| CN101763051B (en) * | 2008-12-24 | 2012-07-25 | 台达电子工业股份有限公司 | Monitor Machine Method |
| CN102693151A (en) * | 2011-03-23 | 2012-09-26 | 鸿富锦精密工业（深圳）有限公司 | Computer and data downloading method thereof |
| CN102263818A (en) * | 2011-07-07 | 2011-11-30 | 北京飞杰信息技术有限公司 | Method for storing and reading file data, and apparatus thereof |
| CN102929791A (en) * | 2012-10-25 | 2013-02-13 | 浪潮电子信息产业股份有限公司 | Cluster storage system expansion method based on identical expansion module |
| CN105471927A (en) * | 2014-05-30 | 2016-04-06 | 华为技术有限公司 | Method for processing service routing message, and service node |
| CN105471927B (en) * | 2014-05-30 | 2019-01-08 | 华为技术有限公司 | The method and service node that a kind of pair of business routing message is handled |
| CN104796774A (en) * | 2015-03-23 | 2015-07-22 | 四川长虹电器股份有限公司 | Across-device application integrating method and system |
| CN108551477A (en) * | 2018-03-28 | 2018-09-18 | 深圳市网心科技有限公司 | Data transmission channel establishment system, network storage device, server and method |
| CN117376289A (en) * | 2023-10-11 | 2024-01-09 | 哈尔滨工业大学 | Network disk data scheduling method for road monitoring data usage application |
| CN117376289B (en) * | 2023-10-11 | 2024-04-12 | 哈尔滨工业大学 | A network disk data scheduling method for road monitoring data usage application |
| CN119127798A (en) * | 2024-11-13 | 2024-12-13 | 杭州浩联智能科技有限公司 | A file storage method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1628296A (en) | System and method for efficient handling of network data | |
| CN1311326C (en) | Methods and apparatus for implementing virtualization of storage within a storage area network through a virtual enclosure | |
| CN1165147C (en) | Wireless communication system and data processing method thereof | |
| US6389432B1 (en) | Intelligent virtual volume access | |
| CN1206600C (en) | Full distribution type aggregation network servicer system | |
| CN1674538A (en) | Network storing system based on local network and its method for reading and writing data | |
| CN1599319A (en) | Method, system, and program for managing data transmission through a network | |
| CN1267418A (en) | Networking systems | |
| CN1852318A (en) | Distributed multi-stage buffer storage system suitable for object network storage | |
| CN1736066A (en) | State engine for data processor | |
| CN101039212A (en) | Fast data storage method | |
| CN1728665A (en) | A scalable object-based storage system and its control method | |
| CN1643872A (en) | Caching streaming data | |
| CN101030946A (en) | Method and system for realizing data service | |
| CN102439574A (en) | Data replacement method in system cache and multi-core communication processor | |
| CN1791020A (en) | Method and system for differentiating data queuing prioritization in a multi-tiered network | |
| CN1949203A (en) | Architecture of interface target machine for miniature computer system and data transmitting method | |
| CN101047580A (en) | Method for setting point-to-point data channel | |
| CN115509644B (en) | Computing power unloading method, device, electronic device and storage medium | |
| CN1791019A (en) | System and method for request priority transfer across nodes in a multi-tier data processing system network | |
| CN1209714C (en) | Server system based on network storage and expandable system structure and its method | |
| CN1753406A (en) | An IP storage control method and device based on iSCSI protocol | |
| CN1220950C (en) | Controller for outer multi-channel network disc array and its protocol fitting method | |
| CN1228726C (en) | Massive memory system based on multi-channel memory equipment and its control method | |
| CN1281037C (en) | Group transmitting system with effective grouping managing unit and operating method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |