CN101984638B - Storage agent system used for cross-IPv4 and IPv6-network processing and method thereof - Google Patents
Storage agent system used for cross-IPv4 and IPv6-network processing and method thereof Download PDFInfo
- Publication number
- CN101984638B CN101984638B CN 201010538475 CN201010538475A CN101984638B CN 101984638 B CN101984638 B CN 101984638B CN 201010538475 CN201010538475 CN 201010538475 CN 201010538475 A CN201010538475 A CN 201010538475A CN 101984638 B CN101984638 B CN 101984638B
- Authority
- CN
- China
- Prior art keywords
- ipv6
- ipv4
- node
- destination node
- packet
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The invention relates to a storage agent system used for cross-IPv4 and IPv6-network processing, comprising a main control unit, an IPv6 input buffer, an IPv4 input buffer, an IPv4 output buffer, an IPv6 output buffer, a PDU analyzer, a PDU constructor, a command analytic unit, a data processing unit, a port mapping tank, a target node registration management module and a communication manager. The method includes that a storage request and storage data packet from an initiator node in the IPv6 network are subject to analyzing, mapping and processing by virtue of the system and are then sent to corresponding storage target node; and a storage response packet and data packet which are sent to the IPv6 node by the target node are subject to buffering and conversion processing by virtue of the system and are then forwarded to the IPv6 initiator node. The invention provides intermediate conversion and expedited forwarding functions for communication and data transmission of IPv6 storage user and IPv4 storage system, and efficiency that an IPv6 uses storage service is improved.
Description
Technical field
The present invention relates to Network storage technology, relate in particular to IPv6 transition period for the treatment of the storage agent system and method across IPv4 and IPv6 network storage service.
Background technology
Along with the fast development of Internet and applying of IPv6 technology, IPv6 and existing IPv4 network integration degree are more and more higher.But the infrastructure of IPv4 network is very extensive, can not accomplish in one move from the IPv4 network transition to IPv6.The IPv6 network interoperability working group (v6ops) of IETF also points out, network can not occur on a large scale, widely, whole migration, transition, IPv4 and IPv6 technology the following more than ten years even in the longer time with long-term co-existence and interoperability.Transition period, IPv6 and IPv4 network were mainly interconnected by modes such as two stacks, tunnel, protocol conversions, that is to say that IPv4 is not by independent pure IPv6 link with the idea that is connected of IPv6 network, transmitted but use original IPv4 link.Stores service is as a kind of valuable source equipment on the other hand, and ISP generally can not dispose many places in network, but concentrates on certain zone.In view of above 2 points, there are the IPv6 of storage demand, IPv4 node all to have competition to the access of stores service equipment at aspects such as the network bandwidth, storage resources, thereby cause performance bottleneck.And research and the solution of current IP v6, IPv4 storage system addressing means aspect relatively lack, and can't satisfy the convenience requirement that the user uses, safeguards storage system.
From the IPv6 agreement, one of IPv6 network design thought is exactly to improve network transmission efficiency, so the packet header length of IPv6 packet shortens, its load can much larger than the IPv4 packet, be specially adapted to large throughput data access services, as application such as Streaming Medias; And using present situation from present IPv6, IPv6 uses many really in the large throughput data access service such as Streaming Media.The specific QoS requirement of service guarantees that this class business need provides requires storage system to support concurrent application as much as possible simultaneously, therefore storage system is produced larger pressure.
The interconnection technique that the network storage is at present adopted mainly contains optical-fibre channel and TCP/IP.IP-based iSCSI(Internet SCSI) agreement is one of IP storage network protocol, and it combines two kinds of industrial standard SCSI and the TCP/IP of extensive use, and the Connection Service of object data block level is provided for memory device.Because the iSCSI agreement can solve the compatibility that exists in the network storage based on optical-fibre channel and transmission range problem and can provide relatively inexpensive solution to receive extensive welcome.
Although the IP network storage based on iSCSI can realize that stores service is to the direct expansion of IPv6 network by in storage server, network layer being increased the IPv6 agreement, but because the modes such as two stacks, tunnel, protocol conversion are used in present IPv6 access mostly, in fact ipv6 traffic uses the IPv4 link transmission, therefore this directly interconnected extension storage method of service can not be brought into play the IPv6 protocol characteristic to greatest extent, and can cause that network traffics are doubled and redoubled, storage resources due to lack unified management, optimization reduces service efficiency.
How to take full advantage of IP6 protocol function and iSCSI Network storage technology characteristics, and the load of reduce IPv4 existing network becomes the key that solves IPv6 network storage issue in transition period.Storage agent in this paper system is a kind of effective solution.
Application number be 200580015737.X, name be called " for the treatment of the method and apparatus across the IPV4 DNS PTR of IPV4 and IPV6 network inquiry " Patent Application Publication for the treatment of the method and apparatus across the IPV4 DNS PTR inquiry of IPV4 and IPV6 network, but do not relate to the storage means of processing across IPV4 and IPV6 network.Application number is 200910242354.6, for the Patent Application Publication of " supporting the service control method of extendible IPv 6 access under the IPv4 environment " service control method of extendible IPv 6 access under the IPv4 environment, but do not relate to network storage aspect.
Summary of the invention
The purpose of this invention is to provide a kind of under IPv6 IPv4/IPv6 in transition period and dis environment, do not change under the condition of existing network configuration, for the communications and data transmission of IPv6 storage user and IPv4 storage system provides intermediate conversion and expedited forwarding function, raising IPv6 user use stores service efficient for the treatment of the storage agent system and method across IPv4 and IPv6 network.
the present invention realizes that the technical scheme of above-mentioned purpose is: a kind of for the treatment of the storage agent system across IPv4 and IPv6 network, comprise: main control unit, IPv6 inputs buffering, IPv4 inputs buffering, IPv4 exports buffering, IPv6 exports buffering, the PDU resolver, the PDU constructor, command resolution unit, data processing unit, the port mapping pond, destination node registration management module and contact manager, described contact manager is by the task management unit, the connection management unit, session management unit, IPv4 end text parameter negotiation information table, IPv6 end text parameter negotiation information table forms.
The present invention realizes in above-mentioned purpose for the treatment of the technical scheme across the storage agent systems approach of IPv4 and IPv6 network is: a kind of for the treatment of the storage agent method across IPv4 and IPv6 network, adopt and describedly process for the treatment of the storage agent system across IPv4 and IPv6 network, comprise the following steps: (1) initialization procedure: destination node registration management module is to the initialization of storage essential information, and contact manager is to the telecommunication management information initializing;
(2) main control unit is received packet and the packet that receives is judged whether it is the IPv6 packet, if IPv6 packet, query aim node information on services registration table and initiator node communication information table, judge whether the initiator node from IPv6, if, put into IPv6 input buffering, otherwise abandon; If the IPv4 packet judges whether the destination node from IPv4, if so, put into IPv4 input buffering, otherwise abandon;
(3) the PDU resolver carries out the packet parsing to the packet that receives after processing through step (2), by packet internal structure judgement type of data packet; According to iSCSI type of message in packet, give respectively command resolution unit or data processing unit and process;
(4) data of processing through step (3) are changed, are configured to new packet through the PDU constructor again, put into corresponding IPv4 output buffer or IPv6 output buffer, wait to be sent;
(5) according to the message communication situation, revise telecommunication management relative recording table by contact manager; Get back to step (2), periodic duty.
Judgement type of data packet in described step (3) specifically refers to:
A checks the telecommunication management state table if based on state message and the command message of iSCSI, further carry out command analysis, and whether judgement order, state be legal, if non-rule is carried out mistake processing;
B checks data validity if based on the data message of iSCSI, check the data transmit-receive record, and invalid data is carried out mistake processing; To the valid data message, carry out Data Division and encapsulation process according to the consultation parameter table.
Destination node information on services registration table in step (2) and initiator node communication information table specifically comprise each field shown in following table:
1. destination node information on services registration table comprises: the IPv4 address of destination node sequence number, destination node name, destination node, destination node communication port numbers, interface group mark;
2. initiator node communication information table comprises: the IPv6 address of initiator node sequence number, initiator node name, initiator node, initiator node communication port numbers, initiator node session id.
5. according to claim 3 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: described telecommunication management state table comprises: duty mapping table, linkage record table, conversation recording table,
(a), the duty mapping table comprises: task number, record the task flagging of initiator node, task order type, read/write data LUN, affiliated hyphen, affiliated session number, task status;
(b), the linkage record table comprises: connect sequence number, the IPv6 address of initiator node, the port numbers of initiator node, the IPv4 address of destination node, the port numbers of destination node, the mapped port that storage agent provides number, affiliated conversation index, task status;
(c), the conversation recording table comprises: conversation index, record the initiator node name, initiator node end session number, record object namespace node, destination node end session number, the port set mark of destination node end, session status.
By each communication tuple between IPv6 initiator node and IPv4 destination node: IPv6 initiator node IPv6 address, IPv6 initiator node port numbers, destination node IP address, destination node port numbers, the mapped port in corresponding ports mapping pond, the initiator node and the destination node that carry out network service in network layer binding are session establishment mapping relations one to one between initiator node and destination node with the network port.
By each Nexus tuple between IPv6 initiator node and IPv4 destination node: IPv6 initiator node name, initiator node end session number, destination node name, port set mark, both sides' interface of a fixed iSCSI session connection; Utilize the Nexus tuple to realize related each session both sides's of iSCSI layer mapping.
Described contact manager provides record, inquiry and update functions to IPv6 initiator node, storage agent system and IPv4 destination node three's communication process, mainly realizes by aspects such as the management of text parameter negotiation information, duty mapping management, connection management, session managements.
The invention has the beneficial effects as follows: the present invention is in IPv6 subnet and the access of the interconnected network boundary of the IPv4 backbone network storage agent system based on iSCSI, and this system supports IPv4/IPv6 dual stack.After resolving, shine upon and process by storage agent with the storage packet from the storage resource request of IPv6 network internal initiator node, send to corresponding storage destination node; Destination node sends to the storage response packet of IPv6 node and packet also to cushion, be transmitted to the IPv6 initiator node after conversion process by storage agent.By the storage agent system, provide the storage system of stores service needn't do any change and just can provide simultaneously unified stores service for IP4 and IPv6 network node, protect as far as possible the existed system investment, take full advantage of existing equipments resources; And all regard as based on IPv4 all storage users, thus simplified transition period IPv4/IPv6 and dis environment under framework and the task management complexity of storage service system.Take full advantage of IPv6 protocol function and iSCSI Network storage technology characteristics, guaranteed efficient access and the operation of IPv6 storage user to the IPv4 memory device, process and the application responds ability for the storage user of IPv6 subnet provides high performance data, guarantee the efficient operation of types of applications system and database.Storage agent provides transparent exchanging visit mediating function for IPv6 stores user and IPv4 storage server and memory device, effectively reduce the resource consumption that IPv6 storage user is used for storage packet encapsulation, transmission and reception, simultaneously reduce the offered load that brings because of IPv6 storage services in a large number of IPv4 storage networking.
To the later stage of IPv6 transition, IPv6 will become backbone network, stores service is mostly based on IPv6, this moment, storage agent can be used as the middle bridge that isolated IPv4 network storage user conducts interviews to IPv6 storage target, make the storage target it seems that all users are IPv6 users and unified stores service is provided, simplified equally storage system framework that the transition later stage takes as the leading factor with IPv6 and the complexity of task management.
Description of drawings
Fig. 1 is network application environment of the present invention;
Fig. 2 is storage agent internal system structure chart of the present invention;
Fig. 3 is the procotol correspondence mapping graph of system of the present invention;
Fig. 4 is the entry stage traffic diagram of storage agent of the present invention system;
Fig. 5 is the write operation traffic diagram of storage agent of the present invention system;
Fig. 6 is the read operation traffic diagram of the storage agent system that the present invention relates to;
Fig. 7 is the workflow diagram that system of the present invention and system relate to;
Fig. 8 is the network application environment that another example of the present invention relates to.
Embodiment
as shown in Figure 2, of the present invention for the treatment of across IPv4 and IPv6 network, internal structure based on the storage agent 205 of the storage agent system of iSCSI storage service comprises: main control unit 100, IPv6 input buffering 101, IPv4 input buffering 105, IPv4 output buffering 103, IPv6 output buffering 108, PDU resolver 104, PDU constructor 107, command resolution unit 102, data processing unit 106, port mapping pond 109, destination node registration management module 110 and contact manager 112, described contact manager 112 is by task management unit 111, connection management unit 113, session management unit 114, IPv4 end text parameter negotiation information table 115, IPv6 end text parameter negotiation information table 116 forms.
As shown in Fig. 1,3,4,5,6,7, a kind of for the treatment of across IPv4 and IPv6 network, based on the storage agent method of iSCSI storage service, adopt and describedly process for the treatment of the storage agent system across IPv4 and IPv6 network, comprise the following steps:
(1) initialization procedure: the initialization of 110 pairs of storage essential informations of destination node registration management module, 112 pairs of telecommunication management information initializings of contact manager;
(2) main control unit 100 is received packet and the packet that receives is judged whether it is the IPv6 packet, if IPv6 packet, query aim node information on services registration table and initiator node communication information table, judge whether the initiator node from IPv6, if, put into IPv6 input buffering, otherwise abandon; If the IPv4 packet judges whether the destination node from IPv4, if so, put into IPv4 input buffering, otherwise abandon;
(3) packet that receives after 104 pairs of PDU resolvers are processed through step (2) carries out packet resolves, by packet internal structure judgement type of data packet; According to iSCSI type of message in packet, give respectively command resolution unit 102 or data processing unit 106 processing;
(4) data of processing through step (3) are changed, are configured to new packet through PDU constructor 107 again, put into corresponding IPv4 output buffer 103 or IPv6 output buffer 108, wait to be sent;
(5) according to the message communication situation, revise telecommunication management relative recording table by contact manager 112; Get back to step (2), periodic duty.
Judgement type of data packet in described step (3) specifically refers to:
A checks the telecommunication management state table if based on state message and the command message of iSCSI, further carry out command analysis, and whether judgement order, state be legal, if non-rule is carried out mistake processing;
B checks data validity if based on the data message of iSCSI, check the data transmit-receive record, and invalid data is carried out mistake processing; To the valid data message, carry out Data Division and encapsulation process according to the consultation parameter table.
Destination node information on services registration table in step (2) and initiator node communication information table specifically comprise each field shown in following table:
1. destination node information on services registration table comprises: the IPv4 address of destination node sequence number, destination node name, destination node, destination node communication port numbers, interface group mark;
2. initiator node communication information table comprises: the IPv6 address of initiator node sequence number, initiator node name, initiator node, initiator node communication port numbers, initiator node session id.
Described telecommunication management state table comprises: duty mapping table, linkage record table, conversation recording table, and concrete field and being illustrated as:
(a), the duty mapping table comprises: task number, record the task flagging of initiator node, task order type, read/write data LUN, affiliated hyphen, affiliated session number, task status;
(b), the linkage record table comprises: connect sequence number, the IPv6 address of initiator node, the port numbers of initiator node, the IPv4 address of destination node, the port numbers of destination node, the mapped port that storage agent provides number, affiliated conversation index, task status;
(c), the conversation recording table comprises: conversation index, record the initiator node name, initiator node end session number, record object namespace node, destination node end session number, the port set mark of destination node end, session status.
As shown in Figure 1, by each communication tuple between IPv6 initiator node and IPv4 destination node: IPv6 initiator node IPv6 address, IPv6 initiator node port numbers, destination node IP address, destination node port numbers, the mapped port in corresponding ports mapping pond 109, the initiator node and the destination node that carry out network service in network layer binding are session establishment mapping relations one to one between initiator node and destination node with the network port.
By each Nexus tuple between IPv6 initiator node and IPv4 destination node: IPv6 initiator node name, initiator node end session number, destination node name, port set mark, both sides' interface of a fixed iSCSI session connection; Utilize the Nexus tuple to realize related each session both sides's of iSCSI layer mapping.
112 pairs of IPv6 initiator node of described contact manager, storage agent system and IPv4 destination node three's communication process provides record, inquiry and update functions, mainly realizes by aspects such as the management of text parameter negotiation information, duty mapping management, connection management, session managements.
Network application environment of the present invention is: backbone network and storage area network be all based on the IPv4 agreement, and backbone network provides access take IPv4/IPv6 dual stack or tunnel style as the IPv6 sub-network.IPv6 storage user's initiator node (Initiator), storage agent and storage destination node (Target) are all supported the iSCSI agreement.
At the storage agent 205 of the interconnected network boundary access of IPv6 subnet 201 and IPv4 backbone network 401 based on iSCSI, storage agent 205 is supported IPv4/IPv6 dual stack.From the storage resource request of the inner initiator node 202 of IPv6 network 201 or 203 and storage packet by storage agent 205 resolve, mapping and process after, send to the respective stored server destination node 502 or 503 of appointment; The storage response packet that storage server destination node 502 or 503 sends to and packet are also by being transmitted to former IPv6 initiator node 202 or 203 after storage agent 205 bufferings, mapping, processing.And still communicate by letter with the destination node 502 or 503 of storage area 501 by IPv4 backbone network 401 from storage resource request and the packet of IPv4 subnet 301, be not subjected to the impact of IPv6 access.A mapped port correspondence of storage agent the corresponding relation of a session between IPv6 initiator node port and destination node port.
A plurality of functional modules that above-mentioned storage agent 205 is comprised are described as follows:
Main control unit 100: the storage agent flow process is carried out control and management.
IPv6 input buffering 101: be used for temporarily depositing the iSCSI protocol massages from the IPv6 initiator node.
IPv6 output buffering 103: be used for temporarily depositing the iSCSI protocol massages that to issue the IPv6 initiator node.
IPv4 input buffering 105: be used for temporarily depositing the iSCSI protocol massages from the IPv4 destination node.
IPv4 input buffering 108: be used for temporarily depositing the iSCSI protocol massages that to issue the IPv4 destination node.
Command resolution unit 102: classification message commands out after resolving by PDU resolver 104 is carried out the context determination analysis, and make suitable reply process according to the iSCSI agreement.
Data processing unit 106: classification message data out after resolving by PDU resolver 104 is carried out analytical calculation, split or union operation according to the data message length limit value of consulting, and deposit new message in output buffer.
Port mapping pond 109: be session establishment mapping relations one to one between initiator node and destination node with the port of storage agent 205.
Destination node registration management module 110: the storing service information that the Collection and preservation destination node provides, for inquiry.
Contact manager 112: the communication process to IPv6 initiator node, storage agent and IPv4 destination node three provides record, inquiry and update functions, mainly realizes by aspects such as the management of text parameter negotiation information, duty mapping management, connection management, session managements.
System specifically describes as follows to above-mentioned storage agent.
The storage basic information management
Form
Destination node information on services registration table
| Target-ID | The destination node sequence number |
| Target-Name | The destination node name |
| IP-Address | The IPv4 address of Target |
| Port | The destination node communication port numbers |
| Portal-group-tag | The interface group mark |
| Initiator-ID | The initiator node sequence number |
| Initiator-Name | The initiator node name |
| IPv6-Address | The IPv6 address of initiator node |
| Port | The initiator node communication port numbers |
| ISID | The initiator node session id |
Each communication tuple between initiator node and destination node: initiator node IPv6 address, initiator node port numbers, destination node IP address, destination node port numbers, corresponding stored agency's a mapped port is bound the both sides' main frame that carries out network service in network layer.
Contact manager
Form
The duty mapping table
| TaskID | Task number |
| Initiator_TaskTag | Record the task flagging (ITT) of initiator node |
| Command | The task order type |
| LUN | The read/write data LUN |
| ConnectionID | Affiliated hyphen |
| SessionID | Affiliated session number |
| Task_Status | Task status |
A linkage record table is safeguarded in connection management unit 113, and the storage agent mappings port realizes connecting both sides' binding, concrete field and being illustrated as:
Form
The linkage record table
| ConnectionID | Connect sequence number |
| Initiator_IPv6_Address | The IPv6 address of initiator node |
| Initiator_Port | The port numbers of initiator node |
| Target_IPv4_Address | The IPv4 address of destination node |
| Target _Port | The port numbers of destination node |
| Agent_map_Port | The mapped port that storage agent provides number |
| SessionID | Affiliated conversation index |
| Connection_Status | Task status |
Form
The conversation recording table
| SessionID | Conversation index |
| Initiator-Name | Record the initiator node name |
| ISID | Initiator node end session number |
| Target-Name | The record object namespace node |
| TSIH | Destination node end session number |
| Target-Portal-Group-Tag | The port set mark of destination node end |
| Session_Status | Session status |
Each Nexus tuple between initiator node and destination node: initiator node name, ISID, destination node name, port set mark, both sides' interface of an iSCSI session connection of binding.The Nexus tuple of utilizing storage agent realizes related each session both sides's of iSCSI layer mapping.
Entry stage
For the convenience of describing, the below is described take communicating by letter as example between IPv6 initiator node 202 and destination node 502.
IPv6 initiator node 202 is obtained current name and the corresponding interface group mark that the iSCSI destination node 502 of stores service is provided by the destination node information on services registration table in inquiry storage agent 205 before bringing into use stores service.
Pass through network monitoring, storage agent 205 will receive the memory access logging request that is sent to the corresponding name of destination node 502 and interface by IPv6 initiator node 202, the source address of this message is the IPv6 address of initiator node 202, destination address is the IPv6 address of storage agent 205, and has stipulated initiator node name, port numbers, destination node name and memory interface group echo.Storage agent 205 is put into IPv6 input buffering 101 with this message and is waited in line to process.
In order to consult the parameters configuration with storage system, storage agent 205 initiatively sends to destination node 502 message of consulting order with text parameter, destination node 502 can be replied the parameter value that proposes to consult, and final storage agent 205 is saved in IPv4 end text parameter negotiation information table to the negotiation result of determining.
IPv6 initiator node 202 also can send the message of consulting order with text parameter to storage agent 205 in order to consult the parameters configuration with destination node 502, and storage agent 205 can provide the parameter value that is applicable to the IPv6 agreement to hold consultation with it.Can no longer be subjected to the IPv4 protocols limit with the setting of the key values such as higher limit (MaxOutstandingR2T) of R2T as the maximum length (FirstBurstLength) of the maximum data segment length (MaxRecvDataSegmentLength) of receive data, the maximum length (MaxBurstLength) of request class data, unsolicited class data, default waiting time (DefaultTime2Wait), each task, but satisfy the IPv6 requirement as far as possible, with the protocol function advantage of performance IPv6.Consult to finish, storage agent 205 is saved in IPv6 end text parameter negotiation information table to the text parameter negotiation result of IPv6 end.
The full work stage
After IPv6 initiator node 202 received that login is replied, storage both sides' communication entered the full work stage by entry stage, and IPv6 initiator node 202 can send the reading and writing data command to stores service equipment as required.
(1) write order communication process
When IPv6 initiator node 202 need to be write storage during data, can send an iSCSI write order and write in the first batch data to the IPv6 address of storage agent 205.Comprise initiator node task flagging (ITT), LUN (LUN), LBA (LBA) and total length in write order.Writing in the first batch the important traffic parameters such as maximum length of data in communication process carries out according to entry stage storage agent 205 and the parameter value that IPv6 initiator node 202 is consulted.After storage agent 205 is received this message, on the one hand, storage agent 205 is after receiving the writing data command and first batch of output data and carry out command analysis of IPv6 initiator node 202, task management unit 111 in the contact manager 112 adds a record, upgrade session management unit 114 and connection management unit 113 according to connecting sequence number, conversation index, and send a R2T(Ready-to-Transfer by PDU constructor 107 to IPv6 initiator node 202) response message, show and carry out the preparation that receives the next group data.On the other hand, resolving, write order is delivered after judgement command resolution unit 102 by PDU resolver 104 processes, will write in the first batch data and be given to data processing unit 106.Data processing unit 106 is repacked after in IPv4 end text parameter negotiation information table, the corresponding entry of destination node 502 provides the consultation parameter value that data are split.The data block of PDU constructor 107 after with write order and Reseal be configured to take the IPv4 address of storage agent 205 to be the write order message of destination address as the IPv4 address of source address, destination node 502, puts into IPv4 output buffering etc. to be sent.Generally much larger than IPv4 packet maximum length, so data processing unit 106 is always the less data message of length with Data Division entrained in the IPv6 packet due to the maximum data packet length of IPv6 agreement regulation.After storage agent 205 receives that the R2T with ITT, buffer offset and total length that sends from destination node 502 replys, will continue to send data remaining to be exported.So circulation is until send the state confirmation response message that all data are received destination node.Task finishes, and storage agent upgrades this task record state for completing from the duty mapping table.
(2) read command communication process
When the IPv6 initiator node need to be from the destination node read data, at first the IPv6 address to storage agent 205 sent a read command.Comprise initiator node task flagging (ITT), LUN (LUN), LBA (LBA) and total length in read command.After storage agent 205 received that command analysis is carried out in this instruction, the task management unit 111 in the contact manager 112 added a record, set up related with session management unit 114 and connection management unit 113; Then reequip message, send to destination node IPv4 address.Destination node receive orders and get ready after begin to send the reading data piece of asking.205 IPv4 packet bufferings of receiving of storage agent are to the IPv4 buffering area, then re-assembly, be configured to take the IPv6 address of storage agent 205 to be the iSCSI reading data message of destination address as the IPv6 address of source address, initiator node according to the corresponding consultation parameter value of IPv6 initiator node in IPv6 end text parameter negotiation information table, put into the IPv6 output buffer and queue up and send.So constantly circulation receives that state replys after all data that will read until receive.Task finishes, and storage agent 205 upgrades this task status from the duty mapping table.
Workflow:
(1) initialization procedure: destination node registration management module is to the initialization of storage essential information, and contact manager is to the telecommunication management information initializing.
(2) storage agent grid interface judges whether it is the IPv6 packet to the packet that receives, if IPv6 packet, query aim node information on services registration table and initiator node communication information table, judge whether the initiator node from IPv6, if, put into IPv6 input buffering, otherwise abandon; If the IPv4 packet judges whether the destination node from IPv4, if so, put into IPv4 input buffering, otherwise abandon;
(3) the PDU resolver carries out the packet parsing to the packet of receiving, by packet internal structure judgement type of data packet; According to iSCSI type of message in packet, give respectively command resolution unit or data processing unit and process;
(4) the PDU constructor is changed, is configured to new packet, puts into corresponding IPv4 or IPv6 output buffer, waits to be sent;
(5) according to the message communication situation, revise telecommunication management relative recording table by contact manager; Get back to step (2), periodic duty.
In concrete enforcement, storage agent of the present invention system can be according to the network environment difference and respective change.At the IPv6 transition initial stage, take the IPv4 network as main, IPv6 is as indivedual subnets accesses, and stores service is based on the IPv4 agreement, so storage agent is that IPv6 user accesses IPv4 stores service equipment and offers help; To the IPv6 transition middle and later periods, take the IPv6 network as main, IPv4 is as indivedual subnets accesses, and stores service is based on the IPv6 agreement, so storage agent is that IPv4 user accesses IPv6 stores service equipment and offers help.
As shown in Figure 8, be in the interconnected network boundary of IPv4 subnet 601 and IPv6 backbone network 701 based on the storage agent system 605 of iSCSI.From the storage resource request of the inner initiator node 602 of IPv4 network 601 and storage packet by storage agent system 605 resolve, mapping and process after, send to the IPv6 storage server destination node 902 of appointment in IPv6 storage networking 901; The storage response packet that storage server destination node 902 sends and packet are transmitted to former IPv4 initiator node 602 by storage agent system 605 bufferings, mapping, after processing.A mapped port correspondence of storage agent system 605 corresponding relation of a session between IPv6 initiator node port and destination node port.
(1) initialization procedure: destination node registration management module is to the initialization of storage essential information, and contact manager is to the telecommunication management information initializing.
(2) storage agent grid interface judges whether it is the IPv4 packet to the packet that receives, if IPv4 packet, query aim node information on services registration table and initiator node communication information table, judge whether the initiator node from IPv4, if, put into IPv4 input buffering, otherwise abandon; If the IPv6 packet judges whether the destination node from IPv6, if so, put into IPv6 input buffering, otherwise abandon;
(3) the PDU resolver carries out the packet parsing to the packet of receiving, by packet internal structure judgement type of data packet; According to iSCSI type of message in packet, give respectively command resolution unit or data processing unit and process;
(4) the PDU constructor is changed, is configured to new packet, puts into corresponding IPv6 or IPv4 output buffer, waits to be sent;
(5) according to the message communication situation, revise telecommunication management relative recording table by contact manager; Get back to step (2), periodic duty.
Can find out from the explanation of above specific embodiment, in early stage of IPv6 transition and latter stage two kinds of network environments, the principle of storage agent of the present invention system and to form structure be identical.
Early stage to the IPv6 transition, realize that by storage agent IPv6 storage user stores communicating by letter and transfer of data between target with IPv4, take full advantage of on the one hand agreement characteristics and the IPv6 network bandwidth of IPv6 in the IPv6 network, structure is with the IPv6 iSCSI storage packet of mass data load, improve efficiency of transmission and bandwidth availability ratio, reduce the IPv6 node at structure and the resources such as CPU, I/O that consume in resolving the iSCSI message; On the other hand, the storage agent system provides the transfer of iSCSI message for IPv6 storage user, realize the bi-directional conversion of iSCSI message between IPv6 and IPv4, and provide buffering for the memorying data flow based on iSCSI, reduce IPv6 message transmission frequency on the IPv4 backbone network, alleviate the load of IPv4 backbone network and the processing complexity of IPv4 storage system.Similarly, in the middle and later periods to the IPv6 transition, realize that by storage agent IPv4 storage user stores communicating by letter and transfer of data between target with IPv6, for IPv4 storage user provides the transfer of iSCSI stored messages and buffering, realize the bi-directional conversion of iSCSI message between IPv4 and IPv6, with loading rate and the efficiency of transmission that takes full advantage of the IPv6 network message, improve network storage efficiency of service, simplify framework and the task management complexity of storage service system in transition period.
Claims (8)
1. one kind for the treatment of the storage agent system across IPv4 and IPv6 network, it is characterized in that: comprising: main control unit (100), IPv6 input buffering (101), IPv4 input buffering (105), IPv4 output buffering (103), IPv6 output buffering (108), PDU resolver (104), PDU constructor (107), command resolution unit (102), data processing unit (106), port mapping pond (109), destination node registration management module (110) and contact manager (112), described contact manager (112) is by task management unit (111), connection management unit (113), session management unit (114), IPv4 holds text parameter negotiation information table (115), IPv6 end text parameter negotiation information table (116) forms,
storage agent (205) as the storage agent system initiatively sends to destination node (502) message of consulting order with text parameter, destination node (502) can be replied the parameter value that proposes to consult, and final storage agent (205) is saved in IPv4 end text parameter negotiation information table to the negotiation result of determining, IPv6 initiator node (202) is in order to consult the parameters configuration with destination node (502), send the message of consulting order with text parameter to storage agent (205), storage agent (205) can provide the parameter value that is applicable to the IPv6 agreement to hold consultation with it, the maximum data segment length (MaxRecvDataSegmentLength) of receive data, the maximum length (MaxBurstLength) of request class data, the maximum length of unsolicited class data (FirstBurstLength), the default waiting time (DefaultTime2Wait), each task can no longer be subjected to the IPv4 protocols limit with the setting of the higher limit (MaxOutstandingR2T) of R2T, but satisfy the IPv6 requirement as far as possible, protocol function advantage with performance IPv6, consult to finish, storage agent (205) is saved in IPv6 end text parameter negotiation information table to the text parameter negotiation result of IPv6 end,
Described PDU resolver (104) carries out packet to the packet of receiving resolves, by packet internal structure judgement type of data packet; According to iSCSI type of message in packet, give respectively command resolution unit or data processing unit and process;
Described command resolution unit (102)
:Classification message commands out after resolving by PDU resolver (104) is carried out the context determination analysis, and make suitable reply process according to the iSCSI agreement;
Described data processing unit (106): classification message data out after resolving by PDU resolver (104) is carried out analytical calculation, split or union operation according to the data message length limit value of consulting, and deposit new message in output buffer;
Described destination node registration management module (110), to the initialization of storage essential information, the storing service information that the Collection and preservation destination node provides is for inquiry;
Described contact manager (112) is to the telecommunication management information initializing, and contact manager (112) provides record, inquiry and update functions to IPv6 initiator node, storage agent and IPv4 destination node three's communication process.
2. one kind for the treatment of the storage agent method across IPv4 and IPv6 network, adopts claimed in claim 1ly to process for the treatment of the storage agent system across IPv4 and IPv6 network, it is characterized in that: comprise the following steps:
(1) initialization procedure: destination node registration management module (110) is to the initialization of storage essential information, and contact manager (112) is to the telecommunication management information initializing;
(2) main control unit (100) is received packet and the packet that receives is judged whether it is the IPv6 packet, if IPv6 packet, query aim node information on services registration table and initiator node communication information table, judge whether the initiator node from IPv6, if, put into IPv6 input buffering, otherwise abandon; If the IPv4 packet judges whether the destination node from IPv4, if so, put into IPv4 input buffering, otherwise abandon;
(3) PDU resolver (104) carries out the packet parsing to the packet that receives after processing through step (2), by packet internal structure judgement type of data packet; According to iSCSI type of message in packet, give respectively command resolution unit (102) or data processing unit (106) and process;
(4) new packet is changed, is configured to the data of processing through step (3) through PDU constructor (107) again, puts into corresponding IPv4 output buffer (103) or IPv6 output buffer (108), waits to be sent;
(5) according to the message communication situation, revise telecommunication management relative recording table by contact manager (112); Get back to step (2), periodic duty.
3. according to claim 2 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: the judgement type of data packet in described step (3) specifically refers to:
A checks the telecommunication management state table if based on state message and the command message of iSCSI, further carry out command analysis, and whether judgement order, state be legal, if non-rule is carried out mistake processing;
B checks data validity if based on the data message of iSCSI, check the data transmit-receive record, and invalid data is carried out mistake processing; To the valid data message, carry out Data Division and encapsulation process according to the consultation parameter table.
4. according to claim 1 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: the destination node information on services registration table in step (2) and initiator node communication information table specifically comprise each field shown in following table:
1. destination node information on services registration table comprises: the IPv4 address of destination node sequence number, destination node name, destination node, destination node communication port numbers, interface group mark;
2. initiator node communication information table comprises: the IPv6 address of initiator node sequence number, initiator node name, initiator node, initiator node communication port numbers, initiator node session id.
5. according to claim 3 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: described telecommunication management state table comprises: duty mapping table, linkage record table, conversation recording table,
(a), the duty mapping table comprises: task number, record the task flagging of initiator node, task order type, read/write data LUN, affiliated hyphen, affiliated session number, task status;
(b), the linkage record table comprises: connect sequence number, the IPv6 address of initiator node, the port numbers of initiator node, the IPv4 address of destination node, the port numbers of destination node, the mapped port that storage agent provides number, affiliated conversation index, task status;
(c), the conversation recording table comprises: conversation index, record the initiator node name, initiator node end session number, record object namespace node, destination node end session number, the port set mark of destination node end, session status.
6. according to claim 3 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: by each communication tuple between IPv6 initiator node and IPv4 destination node: IPv6 initiator node IPv6 address, IPv6 initiator node port numbers, destination node IP address, the destination node port numbers, the mapped port in corresponding ports mapping pond (109), bind initiator node and the destination node that carries out network service in network layer, be session establishment mapping relations one to one between initiator node and destination node with the network port.
7. according to claim 3 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: by each Nexus tuple between IPv6 initiator node and IPv4 destination node: IPv6 initiator node name, initiator node end session number, destination node name, port set mark, both sides' interface of a fixed iSCSI session connection; Utilize the Nexus tuple to realize related each session both sides's of iSCSI layer mapping.
8. according to claim 5 for the treatment of the storage agent method across IPv4 and IPv6 network, it is characterized in that: described contact manager (112) provides record, inquiry and update functions to IPv6 initiator node, storage agent system and IPv4 destination node three's communication process, mainly by the management of text parameter negotiation information, duty mapping management, connection management, session management aspect realization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010538475 CN101984638B (en) | 2010-11-10 | 2010-11-10 | Storage agent system used for cross-IPv4 and IPv6-network processing and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010538475 CN101984638B (en) | 2010-11-10 | 2010-11-10 | Storage agent system used for cross-IPv4 and IPv6-network processing and method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101984638A CN101984638A (en) | 2011-03-09 |
| CN101984638B true CN101984638B (en) | 2013-05-15 |
Family
ID=43641809
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010538475 Expired - Fee Related CN101984638B (en) | 2010-11-10 | 2010-11-10 | Storage agent system used for cross-IPv4 and IPv6-network processing and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101984638B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103023787B (en) * | 2011-09-26 | 2016-01-20 | 百度在线网络技术（北京）有限公司 | Data center systems and device and the method for service is provided |
| CN105450755B (en) * | 2015-12-01 | 2018-12-14 | 深圳市迪菲特科技股份有限公司 | A kind of method and device of the more host network port performances of equilibrium based on ipsan |
| CN108616722B (en) * | 2018-04-18 | 2021-02-26 | 中南大学 | Embedded high-definition video acquisition and data stream transmission system |
| CN109413142B (en) * | 2018-09-07 | 2020-07-31 | 电信科学技术第五研究所有限公司 | Method for realizing iSCSI virtual agent under L inux |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1399742A (en) * | 1999-06-30 | 2003-02-26 | 倾向探测公司 | Method and apparatus for monitoring traffic in network |
| CN101116052A (en) * | 2004-12-21 | 2008-01-30 | 米斯特科技有限公司 | Network interface and firewall device |
| US7594002B1 (en) * | 2003-02-14 | 2009-09-22 | Istor Networks, Inc. | Hardware-accelerated high availability integrated networked storage system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8149711B2 (en) * | 2007-07-25 | 2012-04-03 | Silicon Image, Inc. | Data stream control for network devices |
-
2010
- 2010-11-10 CN CN 201010538475 patent/CN101984638B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1399742A (en) * | 1999-06-30 | 2003-02-26 | 倾向探测公司 | Method and apparatus for monitoring traffic in network |
| US7594002B1 (en) * | 2003-02-14 | 2009-09-22 | Istor Networks, Inc. | Hardware-accelerated high availability integrated networked storage system |
| CN101116052A (en) * | 2004-12-21 | 2008-01-30 | 米斯特科技有限公司 | Network interface and firewall device |
Non-Patent Citations (2)
| Title |
|---|
| J. Satran，K. Meth，C. Sapuntzakis，M. Chadalapaka，E. Zeidner.Internet Small Computer Systems Interface (iSCSI).《RFC3720》.2004, * |
| 杨正富,傅光轩.基于隧道和NAT-PT的新型过渡技术研究与实现.《计算机与现代化》.2009,(第170期), * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101984638A (en) | 2011-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6175867B1 (en) | System and method for managing networks addressed via common network addresses | |
| CN101986666B (en) | Network data transmission method based on virtual network interface and reverse address resolution | |
| CN102170380B (en) | Method and device for accessing outer network from inner network | |
| US8422503B2 (en) | Address translator using address translation information in header area on network layer level and a method therefor | |
| CN101156147A (en) | Method and apparatus for handling ipv4 dns ptr queries across ipv4 and ipv6 networks | |
| US20120082160A1 (en) | Routing using global address pairs | |
| CN101136910B (en) | Network address and protocol translating equipment and application layer gateway equipment | |
| US8892724B1 (en) | Assigning a type of address based on expected port utilization | |
| CN103262505A (en) | Differentiated handling of network traffic using network address translation | |
| CN102209121A (en) | Method and device for intercommunication between Internet protocol version 6 (IPv6) network and Internet protocol version 4 (IPv4) network | |
| CN101316230A (en) | Virtual network connection system, apparatus and medium | |
| CN101141420A (en) | Method and system for performing data communication between private network and public network | |
| CN101984638B (en) | Storage agent system used for cross-IPv4 and IPv6-network processing and method thereof | |
| CN102739809A (en) | DNS64 database, server, system and IPv4/IPv6 communication method | |
| CN102255982A (en) | Internet protocol version 4 (IPv4)/IPv6 translation gateway and translation method | |
| CN103200281A (en) | Method, device and system for accessing intranet server | |
| US8612557B2 (en) | Method for establishing connection between user-network of other technology and domain name system proxy server for controlling the same | |
| WO2003103233A1 (en) | Packet repeating installation, network connection device, packet repeating method, recording medium, program | |
| JP5930449B2 (en) | Data interaction method, apparatus, and system | |
| US9143463B2 (en) | Method for performing protocol translation in a network switch | |
| CN104995610B (en) | Device, system, method, storage medium and computer system for being communicated with the local component for calculating equipment | |
| KR100652958B1 (en) | Method of transmitting data by improvement of translating network address in gateway and system thereof | |
| CN104735073B (en) | IPv4-IPv6 transition protocols dispatching method and device | |
| KR100854681B1 (en) | Gateway and method of interoperating between internet protocol-ubiquitous sensor network and simple network management protocol network | |
| JP3548157B2 (en) | Relay device, address conversion control method, and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130515 Termination date: 20151110 |
|
| EXPY | Termination of patent right or utility model |