CN102143000B

CN102143000B - Method for controlling data recovery, node, control device and disaster-recovery system

Info

Publication number: CN102143000B
Application number: CN201110081138.5A
Authority: CN
Inventors: 姬庆发
Original assignee: ZTE Corp
Current assignee: Nanjing Zhongxing Software Co Ltd
Priority date: 2011-03-31
Filing date: 2011-03-31
Publication date: 2015-06-03
Anticipated expiration: 2031-03-31
Also published as: CN102143000A; WO2012129978A1

Abstract

The invention provides a method for controlling data recovery, a node, a control device and a disaster-recovery system. The control method comprises the steps of: in a process of recovering data, collecting performance indexes of data transmission among regions by the control device; judging congestion level of a network among the regions by the control device according to the performance indexes; and when the congestion level exceeds the preset congestion control threshold, reducing the number of nodes of network recovery data among the regions in the first region by the control device. The invention can reasonably utilize network bandwidth among the regions in a data recovery process, and can avoid that the data transmission efficiency is deteriorated due to network congestion, thereby reducing time required in the disaster-recovery time.

Description

A kind of control method of date restoring, node, control device and disaster tolerance system

Technical field

The present invention relates to Data Disaster-Tolerance Technology field, be specifically related to a kind of control method of date restoring, node, control device and disaster tolerance system.

Background technology

In order to improve the fail safe of data, prior art extensively adopts the disaster tolerance system of redundancy backup.For user data, with communication network technology development and operation demand, at different times in order to realize different business, or complete different purposes, the a lot of network element of operator deployment, and in these network elements, having various user data (User Data), this makes operator more and more be difficult to issue new business, even manages these data also more difficult.For this reason, industry proposes the concept of unification user data model, get up by the data fusion of same user distribution in different network elements, be stored in unified storage of subscriber data network element: user data concentrates (UDC, User Data Convergence) node.Under this implementation, the user data in different network elements/network opens and is shared as user and brings unified Consumer's Experience.Leaving concentratedly and managing of user data is also brought into schedule to the requirement of the fail safe of data thereupon, and therefore UDC considers the factor of the redundancy of data in the design fully and introduces geographical level disaster tolerance.

In geographical level disaster tolerance, the user data deposited in UDC node will be distributed in multiple area and can retain multiple backups of data in each area simultaneously.After a regional equipment occurs damaging and is repaired, all nodes of this area by interlocal network from other normal regional Nodes recovery data, can complete disaster-tolerant recovery process.Please refer to Fig. 1, for the schematic diagram of the disaster-tolerant recovery process of prior art, suppose that damaging appears in all node devices in the first area, each node device in the first area is after being repaired, all will initiate date restoring to the node of bizonal correspondence, thus 6 data recovering process P1 ~ P6 may be there are simultaneously between first, second area.

Although the disaster tolerance of geographical level effectively solves the service disconnection caused because of geographical disaster and device damage, but because the mechanism that introduces interzone date restoring and operator are to the requirement of disaster-tolerant recovery time, the parameter such as bandwidth, time delay of this technology to interzone transmission network proposes very high request, and these require often beyond the available Internet resources of existing network.

Due to the disaster-tolerant recovery process that the restriction of Internet resources needs all nodes carrying out date restoring to initiate in failover procedure in addition all simultaneously, cause fighting for the fierceness of limited bandwidth resource between node, exacerbate the time delay of transfer of data, cause upper layer application in transmitting procedure, occur time-out in data and constantly retransmit, final deterioration data transmission efficiency, further increases the consuming time of disaster-tolerant recovery process.

Summary of the invention

Technical problem to be solved by this invention is to provide a kind of control method of date restoring, node, control device and disaster tolerance system, in order to the interlocal network bandwidth of Appropriate application in data recovery procedure.

For solving the problems of the technologies described above, the invention provides scheme as follows:

A kind of control method of date restoring, the node being applied to the first area recovers, the process of data, wherein, to be connected between the node of areal by local network from bizonal Nodes, connected by interzone network between the node of different regions, described control method comprises:

In the process of described recovery data, the performance index of interzone transfer of data collected by control device;

Described control device, according to described performance index, judges the Congestion Level SPCC of interzone network;

When described Congestion Level SPCC exceeds predetermined congestion control threshold, described control device reduces in described first area simultaneously by the quantity of the node of interzone network recovery data.

Preferably, above-mentioned control method, also comprises:

When described Congestion Level SPCC does not exceed described congestion control threshold, described control device increases in the first area simultaneously by the quantity of the node of interzone network recovery data.

Preferably, in above-mentioned control method,

Comprised by the quantity of the node of interzone network recovery data in described minimizing first area simultaneously:

The data that described control device is corresponding according to each node, become more than one combination of nodes by the node division in the first area, wherein, only comprise the node for storing identical data in each combination of nodes;

Described control device judge whether to exist in described combination of nodes plural, simultaneously by the node of interzone network recovery data;

When there is plural, node simultaneously by interzone network recovery data in described combination of nodes, only allow a node in described combination of nodes by interzone network recovery data, stop other node in described combination of nodes by interzone network recovery data.

Preferably, in above-mentioned control method,

Also comprised by the quantity of the node of interzone network recovery data in described minimizing first area simultaneously:

When passing through no more than 1 of the number of nodes of interzone network recovery data in each described combination of nodes, described control device upgrades the quantity that interzone carries out the quota of the node of data Replica simultaneously, and the quantity of described quota is less than the quantity of the current node by interzone network recovery data in described first area;

Described quota is licensed to the node of respective amount in described first area by described control device;

Described control device only allows the node obtaining quota mandate in described first area to continue through interzone network recovery data, stops not obtain in described first area the node of quota mandate by interzone network recovery data.

Preferably, in above-mentioned control method, also comprise:

There is the node having completed date restoring in described combination of nodes after, the node controlling not yet to complete in described combination of nodes date restoring, by local network, recovers data from the described Nodes having completed date restoring.

Present invention also offers a kind of control device of date restoring, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions, described control device comprises:

Collector unit, in the process of described recovery data, collects the performance index of interzone transfer of data;

First judging unit, for according to described performance index, judges the Congestion Level SPCC of interzone network;

First control unit, for when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduces in the first area simultaneously by the quantity of the node of interzone network recovery data.

Preferably, in above-mentioned control device, also comprise:

Second control unit, for when institute's Congestion Level SPCC does not exceed described congestion control threshold, increases in the first area simultaneously by the quantity of the node of interzone network recovery data.

Preferably, in above-mentioned control device, described first control unit comprises:

Division unit, for the data corresponding according to each node, becomes more than one combination of nodes by the node division in the first area, wherein, only comprises the node for storing identical data in each combination of nodes;

Second judging unit, for judge whether to exist in described combination of nodes plural, simultaneously by the node of interzone network recovery data;

First processing unit, during for there is plural, node simultaneously by interzone network recovery data in described combination of nodes, only allow a node in described combination of nodes by interzone network recovery data, stop other node in described combination of nodes by interzone network recovery data.

Preferably, in above-mentioned control device, described first control unit also comprises:

Updating block, for in each described combination of nodes by no more than 1 of the number of nodes of interzone network recovery data time, upgrade the quantity that interzone carries out the quota of the node of data Replica simultaneously, the quantity of described quota is less than the quantity of the current node by interzone network recovery data in described first area;

Granted unit, for licensing to the node of respective amount in described first area by described quota;

Second processing unit, for only allowing the node obtaining quota mandate in described first area to continue through interzone network recovery data, stops not obtain in described first area the node of quota mandate by interzone network recovery data.

Present invention also offers a kind of disaster tolerance system, comprising:

Be arranged on the node in the first area and be arranged on bizonal node;

Wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions;

Described disaster tolerance system also comprises above-described control device.

Present invention also offers the control method of another kind of data Replica, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions, described control method comprises:

First node in described first area, when recovering data by interzone network from bizonal Nodes, the performance index of the transfer of data in Monitoring Data recovery process;

Described first node sends described performance index to control device, with the described performance index making described control device send based on each node, judge the Congestion Level SPCC of interzone network, and when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduce in described first area simultaneously by the quantity of the node of interzone network recovery data.

Preferably, also comprise in above-mentioned control method:

Described first node receives the first control command that described control device issues, described first control command instruction only allows a node in first node combination to pass through interzone network recovery data, described first node combination is the combination of nodes belonging to described first node, and only comprises in the first area for storing the node of identical data in the combination of described first node;

Described first node, according to described first control command, keeps or stops the process of this node by interzone network recovery data.

Preferably, in above-mentioned control method,

After this node of termination is by the process of interzone network recovery data, described first node judges whether there is the node being carried out date restoring by interzone network in the combination of described first node further;

When not carrying out the node of date restoring by interzone network in described first node combination, described first node carries out to described control device application the license that data are recovered in strange land, and after securing permission, carry out date restoring by described interzone network.

Preferably, also comprise in above-mentioned control method:

Described first node receives the second control command that described control device issues, and described second control command instruction only allows to obtain the node of quota mandate by interzone network recovery data;

After this node of termination is by the process of interzone network recovery data, described first node judges whether there is the node being carried out date restoring by interzone network in the combination of described first node;

When not carrying out the node of date restoring by interzone network in described first node combination, described first node carries out to described control device application the quota that data are recovered in strange land, and after the mandate of acquisition quota, carry out date restoring by described interzone network.

Preferably, in above-mentioned control method

Whether before judging to exist in the combination of described first node and carrying out the node of date restoring by interzone network, described first node judges whether there is the node having completed date restoring in the combination of described first node further;

When there is the node completing date restoring in described first node combination, described first node, by local network, recovers data from the described Nodes having completed date restoring;

When there is not the node having completed date restoring in described first node combination, described first node judges whether there is the node being carried out date restoring by interzone network in the combination of described first node.

Present invention also offers a kind of first node, described first node is arranged in the first area, and described first node comprises:

Monitoring means, for when this node recovers data by interzone network from bizonal Nodes, the performance index of the transfer of data in Monitoring Data recovery process;

Report unit, for sending described performance index to control device, with the described performance index making described control device send based on each node, judge the Congestion Level SPCC of interzone network, and when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduce in the first area simultaneously by the quantity of the node of interzone network recovery data.

Preferably, in above-mentioned first node, also comprise:

First receiving element, for receiving the first control command that described control device issues, described first control command instruction only allows a node in first node combination to pass through interzone network recovery data, described first node combination is the combination of nodes belonging to described first node, and only comprises in the first area for storing the node of identical data in the combination of described first node;

First control unit, for according to described first control command, keeps or stops the process of this node by interzone network recovery data.

Preferably, in above-mentioned first node, also comprise:

First judging unit, for after described first control unit stops the process of this node by interzone network recovery data, judges whether there is the node being carried out date restoring by interzone network in the combination of described first node;

First processing unit, during for judging at described first judging unit not carry out the node of date restoring by interzone network in the combination of described first node, the license that data are recovered in strange land is carried out to described control device application, and after securing permission, carry out date restoring by described interzone network.

Preferably, in above-mentioned first node, also comprise:

Second receiving element, for receiving the second control command that described control device issues, described second control command instruction only allows to obtain the node of quota mandate by interzone network recovery data;

Second control unit, for according to described second control command, keeps or stops the process of this node by interzone network recovery data;

Second judging unit, for after described second control unit stops the process of this node by interzone network recovery data, judges whether there is the node being carried out date restoring by interzone network in the combination of described first node;

Second processing unit, during for judging at described second judging unit not carry out the node of date restoring by interzone network in the combination of described first node, the quota that data are recovered in strange land is carried out to described control device application, and after the mandate of acquisition quota, carry out date restoring by described interzone network.

Preferably, in above-mentioned first node, also comprise:

3rd judging unit, for judging whether there is the node having completed date restoring in the combination of described first node;

3rd processing unit, during for there is the node completing date restoring in described first node combination, by local network, recovers data from the described Nodes having completed date restoring.

As can be seen from the above, the control method of date restoring provided by the invention, node, control device and disaster tolerance system, based on the quantity of congestion state dynamic adjustments interzone recovering process, when interzone network occurs congested, be reduced by the quantity that interzone network carries out the node of date restoring, thus the interlocal network bandwidth of Appropriate application in data recovery procedure, can avoid because network congestion worsens data transmission efficiency, thus reduce disaster-tolerant recovery process required time.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of the disaster-tolerant recovery process of prior art;

The schematic diagram of the control method of a kind of date restoring that Fig. 2 provides for the embodiment of the present invention;

Fig. 3 is the schematic diagram of a kind of date restoring in the embodiment of the present invention;

Fig. 4 is the schematic diagram of another kind of date restoring in the embodiment of the present invention;

The structural representation of the control device that Fig. 5 provides for the embodiment of the present invention;

The detail flowchart of the control method of the date restoring that Fig. 6 provides for the embodiment of the present invention;

Fig. 7 is that in the embodiment of the present invention, when enabling the recovery policy of " whole nodal parallel recovers data from strange land ", node recovers the flow chart of data from strange land;

Fig. 8 is that in the embodiment of the present invention, when enabling the recovery policy of " parallel from strange land recovery data between combination of nodes ", node recovers the flow chart of data from strange land;

Fig. 9 is that embodiment of the present invention node when enabling the recovery policy of " quantity of recovery process concurrent between by quota restrictions combination of nodes " recovers the flow chart of data from strange land.

Embodiment

The present invention, by monitoring data transmission performance in interlocal disaster-tolerant recovery process, reduces the number of processes of the simultaneously concurrent date restoring of interzone when network congestion, thus realizes the Appropriate application to interzone network bandwidth resources.Below with reference to accompanying drawing, by specific embodiment, the present invention is described further.

Embodiment one

The control method of the date restoring described in the embodiment of the present invention, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions.Also comprise a control device in the present embodiment, described control device can both communicate with the node in each area, and described control device can be arranged in any area.

In the present embodiment, interzone network can adopt the public networks such as internet, and local network can adopt the LAN such as Ethernet.Usually, the bandwidth of interzone network is less than the bandwidth of local network, or when having same band, interzone network needs to pay more expenses than local network.Therefore, for same bandwidth, interzone network needs to pay more cost than local network, and interlocal Internet resources are usually more limited.

From the angle of control device, the present embodiment illustrates how the present invention implements.

Please refer to Fig. 2, the control method of a kind of date restoring that the embodiment of the present invention provides, comprises the following steps:

Step 21, in the process of described recovery data, the performance index of interzone transfer of data collected by control device.

Here, collect performance index by control device, concrete collection mode can be that control device periodically goes inquiry to each node, or each node periodically initiatively reports to control device, by the performance index of control device summarizing and reporting.The parameters such as time delay when described performance index can be specifically interzone transfer of data or packet loss.

Step 22, described control device, according to described performance index, judges the Congestion Level SPCC of interzone network.

Here, the Congestion Level SPCC of interzone network comprises and occurs congested and do not occur situations such as congested (comprising from congested recovery normal), can also be divided into the congested of multiple rank further when occurring congested according to the congested order of severity.Wherein judge that the mode of Congestion Level SPCC has multiple:

Such as, can first area and the second area between network normal time, predetermined statistics obtains the time delay range of interzone transfer of data, then, control device is based on the performance index collected, judge whether the time delay of interzone transfer of data has exceeded the maximum (also the basis of maximum in above-mentioned time delay range can increase an extra surplus capacity and obtain a time delay threshold value) of above-mentioned time delay range, when time delay exceedes above-mentioned maximum or time delay threshold value, judge that interzone network occurs congested; Otherwise, judge that interzone network does not occur congested, be in normal condition.

Again such as, can pre-set the threshold value of a packet loss, then, control device, based on the performance index collected, judges whether the packet loss of interzone transfer of data exceedes this threshold value, if so, then judges that interzone network occurs congested; Otherwise, judge that interzone network is normal.

Again such as, time delay and packet loss can also be considered to formulate different determination strategy, when packet loss and the arbitrary parameter of time delay exceed corresponding thresholding, namely judge that the appearance of interzone network is congested.

Judge that the congestion state of interzone network recovers normal, then can adopt the criterion contrary with above judgment mode, repeat no more herein.

Step 23, when described Congestion Level SPCC exceeds predetermined congestion control threshold, described control device reduces in described first area simultaneously by the quantity of the node of interzone network recovery data.

Can find out, the quantity of the present embodiment dynamic adjustments interzone recovering process, when interzone network occurs congested, be reduced by the quantity that interzone network carries out the node of date restoring, thus the interlocal network bandwidth of Appropriate application in data recovery procedure, can avoid because network congestion worsens data transmission efficiency, thus reduce disaster-tolerant recovery process required time.

The control method of the date restoring described in the present embodiment, described interlocal network do not occur congested (as network always normal or network from congested revert to normal) time, can also comprise the following steps:

Step 24, when described Congestion Level SPCC does not exceed described congestion control threshold, described control device increases in the first area simultaneously by the quantity of the node of interzone network recovery data.

Here congestion control threshold can be according to the determined numerical value of network actual application environment, can also be the scope of a numerical value, prescribes a time limit, can perform above-mentioned steps 24 at Congestion Level SPCC lower than the lower of this scope; Prescribe a time limit higher than the upper of this scope at Congestion Level SPCC, above-mentioned steps 23 can be performed.

In addition, because congestion control threshold is a numerical value arranging voluntarily of the present embodiment or scope, what therefore this thresholding can arrange by the present embodiment is lower, namely when network reality does not occur congested, namely start congestion control procedure, perform above-mentioned steps 23, reduce in described first area simultaneously by the quantity of the node of interzone network recovery data, higher bandwidth can be provided, to accelerate the resume speed of these nodes like this for the remaining node by interzone network recovery data.

Can find out, the present embodiment is when described interlocal network recovery is normal, and described control device increases simultaneously by the quantity of the node of interzone network recovery data in the first area, to make full use of inter-zone bandwidth, accelerates data recovery procedure.

Referring to Fig. 3 and Fig. 4, the mode of the minimizing number of nodes that can adopt in above-mentioned steps 23 is described.

Please refer to Fig. 3, partial data comprises data 1 ~ 3 totally three parts, and the first area retains two backups of partial data simultaneously, and wherein, node 11 ~ 13 is respectively used to store data 1 ~ data 3, forms a data backup; Node 14 ~ 16 is respectively used to store data 1 ~ data 3, forms another data backup.Bizonal node 21 ~ 23 is respectively used to store data 1 ~ data 3, forms a data backup.

Suppose that damaging (only illustrate, the present embodiment does not require that damaging all appears in all node devices in the first area, and the present embodiment stands good when damaging appears in part of nodes equipment) all appears in all node devices in the first area here.After these nodes are repaired, will pass through first, second interlocal network, Nodes corresponding from the second area recovers data, to carry out date restoring simultaneously.In the present embodiment, it is congested that control device periodically judges whether interzone network occurs, if it is congested that control device judges that interzone network occurs, so control device can reduce simultaneously by the quantity of the node of interzone network recovery data by following steps:

Step 231, the data that described control device is corresponding according to each node, become more than one combination of nodes by the node division in the first area, wherein, only comprise the node for storing identical data in each combination of nodes.

Such as, in figure 3, the node 11 and node 14 that are used for storing data 1 can be divided into combination of nodes 1, the node 12 and node 15 that are used for storing data 2 are divided into combination of nodes 2, the node 13 and node 16 that are used for storing data 3 are divided into combination of nodes 3.

Step 232, described control device judge whether to exist in described combination of nodes plural, simultaneously by the node of interzone network recovery data.

Here, oneself current state can be reported control device by node, or control device initiatively goes the state of inquiring about each node, thus control device can obtain the state of each node in real time, whether comprise node by interzone network recovery data, whether node completes date restoring etc.

Step 233, when there is plural, node simultaneously by interzone network recovery data in described combination of nodes, only allow a node in described combination of nodes by interzone network recovery data, stop other node in described combination of nodes by interzone network recovery data.

Here, in above-mentioned steps 232, judge that all two nodes in each combination of nodes are at the same time by interzone network recovery data, a node that can stop in step 233 in combination of nodes carries out date restoring by interzone network, and only allows one of them node to carry out date restoring by interzone network.Such as, control device can send one first control command to node, and described first control command instruction only allows a node in first node combination to pass through interzone network recovery data; Like this, by the first above-mentioned control command, the strategy of date restoring between current locale can be informed to each node by control device; Node, after receiving this first control command, judges whether to allow this node by interzone network recovery data, and then determines to continue or stop to carry out date restoring by interzone network.

Fig. 3 shows and only allows a node in combination of nodes by the schematic diagram of a kind of date restoring of interzone network recovery data, now only have P10, P12, P14 tri-processes by interzone network recovery data simultaneously, after combination of nodes inside has 1 node to complete date restoring, by process P11, P13, P15, carry out date restoring by local network.

Like this, the process minimizing half of being carried out date restoring by interzone network that can be simultaneously concurrent, thus date restoring taking the interzone network bandwidth can be greatly reduced, improve the data transmission efficiency of interzone network.

After above-mentioned steps 233, although only have at most 1 node in each combination of nodes by interzone network recovery data, but still there is the still congested possibility of interzone network.By periodically monitoring, control device judges that the appearance of interzone network is congested, and no more than 1 is being judged in each described combination of nodes by the number of nodes of interzone network recovery data by above step 232, now, the present embodiment further by following steps, can also alleviate the Congestion Level SPCC of interzone network:

Step 234, when passing through no more than 1 of the number of nodes of interzone network recovery data in each described combination of nodes, described control device upgrades the quantity that interzone carries out the quota of the node of data Replica simultaneously, and the quantity of described quota is less than the quantity of the current node by interzone network recovery data in described first area.

Here, control device is when finding no more than 1 of the number of nodes by interzone network recovery data in each described combination of nodes, need to upgrade (when determining quota for the first time to the data of quota, can according to network congestion degree, initial quota quantities are set, these initial quota quantities are less than the quantity of combination of nodes), the quota quantities after upgrading are reduced further.Can be different according to the Congestion Level SPCC of network, determine the current number needing to reduce, when Congestion Level SPCC is serious, the number of minimizing is larger; Otherwise when Congestion Level SPCC is not too serious, the number of minimizing is less.

Here, described control device can also by sending one second control command to node, and described second control command instruction only allows to obtain the node of quota mandate by interzone network recovery data; Like this, by the second above-mentioned control command, the strategy of date restoring between current locale can be informed to each node by control device; Node, after receiving this second control command, judges whether this node obtains quota, and then determines to continue or stop to carry out date restoring by interzone network.

Step 235, described quota is licensed to the node of respective amount in described first area by described control device.

Here, concrete authorization can be that control device is directly selected node and authorizes, or by node initiatively quota application, control device, after receiving application, according to current residual quota quantities, determines whether being authorized.

Step 236, described control device only allows the node obtaining quota mandate in described first area to continue through interzone network recovery data, stops not obtain in described first area the node of quota mandate by interzone network recovery data.

Here, control device is by sending a suspended market order to the node of not obtaining the authorization, this node, after receiving this suspended market order, stops carrying out date restoring by interzone network, thus stops not obtain the node of quota mandate by interzone network recovery data in the first area.

A kind of schematic diagram recovering data after Fig. 4 shows quotas administered, supposes that quota quantities are 1, then simultaneously only allow process resumption data between area, be followed successively by P21, P23 and P25, after a upper process whether quota, just starts next process.Local can carry out date restoring by internal network.

By above-mentioned steps, the present embodiment can be alleviated network congestion further, improves the efficiency of transmission of interzone data.

In the present embodiment, control device is according to the node status information collected, when there is the node having completed date restoring in combination of nodes, the node controlling not yet to complete in described combination of nodes date restoring, by local network, recovers data from the described Nodes having completed date restoring.The local network that can make full use of bandwidth resources so more sufficient carries out date restoring, reduces the data volume of interzone date restoring, improves date restoring speed, reduces disaster-tolerant recovery required time.

As can be seen from the above:

The present embodiment is under the condition of interzone network bandwidth abundance, and the preferential recovery policy adopting all nodal parallel, to reach the fastest recovery effects, shortens the time of fault recovery.

When the network bandwidth exists bottleneck, the present embodiment adopts between combination of nodes parallel, and the strategy in batches recovered in combination, greatly reduces taking and fighting for IP bandwidth resources in each node recovery process of interzone IP network bandwidth resource.A node is all had preferentially to complete date restoring and put into operation immediately after having recovered owing to ensure that in the combination of each back end, therefore the failure recovery time of this flow process is identical with the failure recovery time under network sufficiency, but greatly reduces the demand of the interzone network bandwidth.

When the network bandwidth cannot meet above-mentioned two kinds of scenes, the present embodiment by the concurrency between the mode dynamic conditioning combination of nodes of employing quotas administered, thus reduces the conflict of inter-zone bandwidth resource occupation, shortens data restore time.Compared with above-mentioned two kinds of scenes, adopt the recovery management mode of quota can network bandwidth resources between area of application to greatest extent, between efficiency and actual effective bandwidth, reach balance.

Based on the control method of above date restoring, the present embodiment additionally provides a kind of control device of date restoring, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions, please refer to Fig. 5, described control device comprises:

Preferably, above-mentioned control device also comprises:

Concrete, described first control unit specifically comprises:

Further, described first control unit also comprises:

Preferably, described first control unit also comprises:

3rd processing unit, for exist in described combination of nodes complete date restoring node after, the node controlling not yet to complete in described combination of nodes date restoring, by local network, recovers data from the described Nodes having completed date restoring.

The present embodiment also provides a kind of disaster tolerance system, and this disaster tolerance system comprises:

Be arranged on the node in the first area, be arranged on bizonal node and control device as described above;

Wherein, described control device comprises:

Wherein, described first control unit concrete structure as described above, repeats no more herein.

Embodiment two

The present embodiment is from the angle of the node in the first area (for convenience of description, be first node hereinafter referred to as this node) to illustrate how the present invention implements.

The control method of a kind of date restoring that the embodiment of the present invention provides, comprises the following steps:

Step 301, the first node in described first area, when recovering data by interzone network from bizonal Nodes, the performance index of the transfer of data in Monitoring Data recovery process;

Step 302, described first node sends described performance index to control device, with the described performance index making described control device send based on each node, judge the Congestion Level SPCC of interzone network, and when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduce in described first area simultaneously by the quantity of the node of interzone network recovery data.

Above-mentioned performance index, also make described control device when judging that described interzone network does not occur congested based on described performance parameter, increase in the first area simultaneously by the quantity of the node of interzone network recovery data.

Like this, control device can according to the performance index collected, and when interzone network congestion, by reducing in the first area simultaneously by the quantity of the node of interzone network recovery data, improving Congestion Level SPCC, improving the data transmission efficiency of interzone network.

In the present embodiment, the control command that first node can also issue according to control device, the strategy of the interzone date restoring of current employing in acquisition system, and then determine whether pass through interzone network recovery data.Now, the present embodiment can also comprise the following steps:

Step 311, described first node receives the first control command that described control device issues, described first control command instruction only allows a node in first node combination to pass through interzone network recovery data, described first node combination is the combination of nodes belonging to described first node, and only comprises in the first area for storing the node of identical data in the combination of described first node;

Step 312, described first node, according to described first control command, keeps or stops the process of this node by interzone network recovery data.

After above-mentioned steps 312, if do not carried out other node of date restoring in first node combination by interzone network, then first node can also be permitted to control device application, so that by interzone network recovery data after securing permission.Now, the present embodiment can also comprise the following steps:

Step 313, after this node of termination is by the process of interzone network recovery data, described first node judges whether there is the node being carried out date restoring by interzone network in the combination of described first node further.

Step 314, when not carrying out the node of date restoring by interzone network in described first node combination, described first node carries out to described control device application the license that data are recovered in strange land, and after securing permission, carries out date restoring by described interzone network.

Step 315, exist when carrying out the Section Point of date restoring by interzone network in the combination of described first node, described first node can enter wait state, until after Section Point completes date restoring, can recover data by local network from Section Point.Certainly, if inter-zone bandwidth allows, when Section Point completes date restoring and do not carry out the node of date restoring by interzone network in first node combination, first node also can carry out date restoring by interzone network.

When interzone network congestion is more serious, control device can also reduce by quota mode the quantity can being carried out the node of date restoring by interzone network further.Now, the present embodiment can also comprise the following steps:

Step 321, described first node receives the second control command that described control device issues, and described second control command instruction only allows to obtain the node of quota mandate by interzone network recovery data.

Like this, by the second control command, the strategy of current interzone date restoring is informed to each node by control device.Whether each node obtains quota according to self, determines whether continuing interlocal date restoring.

Step 322, if first node judges that self does not obtain quota, then can stop by interzone network recovery data; After this node of termination is by the process of interzone network recovery data, described first node judges whether there is the node being carried out date restoring by interzone network in the combination of described first node.

Step 323, when not carrying out the node of date restoring by interzone network in described first node combination, described first node carries out to described control device application the quota that data are recovered in strange land, and after the mandate of acquisition quota, carries out date restoring by described interzone network.

Be explained above first node quota application to carry out the process of interzone date restoring.

In above-mentioned steps 313 and step 322, whether before judging to exist in the combination of described first node and carrying out the node of date restoring by interzone network, described first node can also judge whether described first node exists the node having completed date restoring in combining further:

If there is the node completing date restoring in described first node combination, then described first node can pass through local network, recovers data, to be reduced by the data volume of interzone Internet Transmission from the described Nodes having completed date restoring;

If when there is not the node having completed date restoring in described first node combination, described first node just judges whether there is the node being carried out date restoring by interzone network in the combination of described first node further.

Based on the control method of above date restoring, the present embodiment additionally provides a kind of first node, and described first node is arranged in the first area, and described first node specifically comprises:

Preferably, the above-mentioned performance parameter reporting unit to report, also makes described control device when judging that described interzone network does not occur congested based on described performance parameter, increases in the first area simultaneously by the quantity of the node of interzone network recovery data.

The control command that first node can also issue according to control device, determines the strategy of date restoring between current locale, and then determines how to carry out date restoring.Such as, by strategy restriction each combination of nodes in carry out the number of nodes of date restoring by interzone network time, above-mentioned first node also comprises:

Again such as, when being carried out quantity (the limiting by authorizing quota) of the combination of nodes of date restoring by interzone network by strategy restriction, above-mentioned first node also comprises:

In order to utilize local network to carry out date restoring as much as possible, above-mentioned first node also comprises:

Above two embodiments describe specific embodiment of the invention from the angle of control device and the angle of first node respectively.In order to help to understand content of the present invention further on the whole, following examples pass through the refinement figure of method overall flow figure and part run, are described specific embodiment of the invention.

Embodiment three

The present embodiment is for UDC system, and the node remaining the first area carries out the process of date restoring from bizonal Nodes.The present embodiment based on the Congestion Level SPCC of interzone network, the strategy of dynamic conditioning interzone date restoring:

The control method of the date restoring described in the embodiment of the present invention, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions.Also comprise a control device in the present embodiment, described control device can both communicate with the node in each area, and described control device can be arranged in any area.The control method of a kind of date restoring that the embodiment of the present invention provides, comprises the following steps:

Step 41, when the process of described recovery data starts, controls all nodes needing to carry out data Replica in the first area, all passes through interzone network and recovers data from the Nodes that the second area is corresponding.That is, when the present embodiment log-on data is recovered, the date restoring strategy that acquiescence adopts all nodes all concurrent from strange land, to reduce data restore time.

Step 42, in the process of described recovery data, the performance index of interzone transfer of data collected by control device, and according to described performance index, judge whether interzone network occurs congested:

As congested in do not occurred, then keep current date restoring strategy, namely adopt the date restoring strategy that all nodes are all concurrent from strange land;

As congested in occurred, then reduce in described first area simultaneously by the quantity of the node of interzone network recovery data, to adopt when the interzone network bandwidth is not enough the interzone date restoring strategy optimized, thus efficiency utilization limited bandwidth conditions, reach good date restoring effect.

After reducing the quantity simultaneously by the node of interzone network recovery data in described first area, the present embodiment still monitors interzone congestion situation, if the congested elimination of interzone, the present embodiment can also increase the quantity by the node of interzone network recovery data further, such as, under the date restoring strategy enabling " parallel from strange land recovery data between combination of nodes " or " quantity of recovery process concurrent between by quota restrictions combination of nodes ", if the congested elimination of interzone, then can increase the process of 1 or 2 concurrent recovery of interzone.

Below in conjunction with Fig. 6, the date restoring strategy of the control method of above-mentioned date restoring and employing thereof is described in further detail.

Fig. 6 is the flow chart selecting suitable recovery policy for network bandwidth performance involved by the present embodiment.Its concrete steps are as follows:

Step 401:UDC system (specifically can be performed by control device) start by set date is to the inspection of disaster-tolerant recovery state, if the first area recovers to need to recover data by IP network from the Nodes in strange land (as the second area) after fault, enter step 402, otherwise process ends.

Step 402:UDC system needs returning to form of judgement first area before initiation recovers flow process: if current locale does not also initiate to recover flow process, then adopt nodal parallel from the recovery policy of strange land recovery data according to design preferences; If the first area has initiated to recover flow process, need to judge the bandwidth performance of current network is to determine adopting which kind of recovery policy further.

Step 403: the first area not yet enters recovery flow process, UDC Operation system setting first area return to form as " start recover ", and the recovery policy adopting whole nodal parallel to recover data from strange land recovers data, i.e. all nodes in the first area all concurrent recovery data from strange land from strange land node.

Step 404: because the first area has started recovery flow process, UDC system needs to collect the performance parameter in user data package transmitting procedure, as time delay and time-out information, as the foundation of follow-up selection recovery policy in recovery process.

Based on performance parameter, step 405:UDC system judges that whether interzone network is congested, and exceed predetermined congestion control threshold, specifically can by checking whether the Delay of current collection has remarkable deterioration with time delay during network idle, or frequently there is message dilivery time-out: if there is not network quality deterioration, then keep existing recovery policy until complete recovery process, otherwise need to perform the adjustment of further recovery policy.

Step 406:UDC system, before the adjustment of execution recovery policy, needs to determine current executed recovery policy:

If the recovery policy of current execution is whole nodal parallel recover data from strange land, then adjustment and recovery strategy is parallel from strange land recovery data between combination of nodes, then enter step 407 to perform, wherein, between combination of nodes, the parallel recovery policy recovering data from strange land refers to that all combination of nodes are all parallel from strange land recovery data, but a node can only be had in each combination of nodes of synchronization to recover data from strange land;

If the recovery policy of current execution is parallel from strange land recovery data between combination of nodes, then adjustment and recovery strategy is: by the quantity of recovery process concurrent between quota restrictions combination of nodes, enter step 409 to perform, wherein, being referred to by the recovery policy of the quantity of recovery process concurrent between quota restrictions combination of nodes to allow in each combination of nodes maximum one to obtain the node of quota and recover data from strange land at synchronization, and the quantity of quota is less than the quantity of combination of nodes;

If the recovery policy of current execution is the quantity by recovery process concurrent between quota restrictions combination of nodes, then needs to consider to reclaim allocated quota by force, to reduce concurrent quantity, reduce taking of inter-zone bandwidth, enter step 411 and perform.

Step 407: all recovery process that termination is performing also enables the parallel recovery policy recovering data from strange land between combination of nodes.

Step 408:UDC system constructing combination of nodes, according to the data stored in node by all for this area node division to different combination of nodes, only have a node to be endowed license in the middle of each combination, can recover data by interzone IP network, residue node all enters wait state.After the node secured permission completes date restoring, data are recovered from this node in inner networking by residue node territory of use.All node datas recover to terminate rear disaster-tolerant recovery flow process to be terminated.Certainly, each node in each combination of nodes also can be allowed here to recover data from strange land successively, but data can only be recovered from strange land by 1 node at most in each combination of nodes simultaneously.

When step 409:UDC systematic evaluation recovery policy is the quantity by recovery process concurrent between quota restrictions combination of nodes, the quota quantities of configuration are licensed to the node of respective amount, in each combination of nodes, 1 node can only be had at most to obtain the authorization simultaneously.The combination of nodes obtaining quota mandate keeps current recovery flow process.

The recovery flow process that the combination of nodes stopping not obtain quota is by force initiated by step 410:UDC system, is forcibly stopped the combination of nodes recovering flow process and will enters wait state, until the combination of nodes having obtained quota completes strange land recovery process release quota.

Step 411: under the bandwidth of extreme networking, the quota quantities of UDC system configuration have exceeded the maximum quantity that between current locale, bandwidth is supported.Now because the Concurrency Access between combination of nodes causes network quality to continue to worsen, UDC system needs startup to optimize the maximum quota quantities of dynamic reduction system support further: each recovery by force quota, and the combination of nodes being recovered quota will stop strange land immediately and recover flow process release bandwidth resources.System is optimized continuing, until system can quota quantities reduce to 1 or be that network quality recovers normal.

Please refer to Fig. 7, Fig. 7 is that in the present embodiment, when enabling the recovery policy of " whole nodal parallel recovers data from strange land ", node recovers the flow chart of data from strange land, and its concrete steps are as follows:

Step 501: under the recovery policy of " whole nodal parallel recovers data from strange land ", UDC system does not make any restriction to the recovery behavior of node, and all nodes all can initiate recovery request to strange land enabled node.The time delay occurred in recovery process interior joint needs collection recovery data transfer procedure and time-out information optimize and revise use for follow-up.

Please refer to Fig. 8, Fig. 8 is that in the present embodiment, when enabling the recovery policy of " parallel from strange land recovery data between combination of nodes ", node recovers the flow chart of data from strange land, and its concrete steps are as follows:

Step 601: when using the parallel recovery policy from strange land recovery data between combination of nodes, first node checks in the combination of nodes of place that before initiating to recover whether exist can as recovering the node of data source, if existed, then recover data by the networking of local from this node, enter step 603 and process; Otherwise enter step 602 and continue process.

Step 602: need when node initiates recovery request to judge whether (store the node of same subscriber data in this area according to planning) in this combination of nodes has existed the node just recovering data from strange land, if there is no, then need recover license from the application of UDC system and recover data from strange land node, enter step 605 and process; Otherwise needs wait for that the node of the joint that secures permission recovers data from this node after completing recovery.Enter step 604 to process.

Step 603: node recovers data by the networking of local from local data source, to complete date restoring.

Step 604: if deposited the node that performing recovery in present combination and this node does not complete the flow process recovering data from strange land, then in combination, other node must enter wait state until this node completes recovery process.

Step 605: because present node combination does not start to recover data from strange land, then any node all can recover the license of data to system application to strange land.

Step 606: the node secured permission can be selected data source and start to recover flow process in strange land.Time delay and the time-out information of collecting transfer of data is needed in recovery process.

Fig. 9 is that in the present embodiment, when enabling the recovery policy of " quantity of recovery process concurrent between by quota restrictions combination of nodes ", node recovers the flow chart of data from strange land, and its concrete steps are as follows:

Step 701: under the recovery policy of the quantity by recovery process concurrent between quota restrictions combination of nodes, first node checks that whether exist in its place combination of nodes can as the node of data source before initiation recovers, if existed, recover data by the networking of local from this node, enter step 703 and process; Otherwise enter step 702 and continue process.

Step 702: need when node initiates recovery request to judge whether (store the combination of nodes of same subscriber data in this area according to planning) in this combination has existed the node just recovering data from strange land, if there is no, then need the quota recovered from UDC system application strange land, enter step 705 to process, otherwise needs wait for that the node of the joint that secures permission recovers data from this node after completing recovery, enter step 704 and process.

Step 703: node recovers data by local networking and completes date restoring from local data source.

Step 704: if there is the node that performs and recover in present combination and this node does not complete and recovers the flow process of data from strange land, then in combination, other node must enter wait state until this node completes recovery process.

Step 705: node needs the quota recovered to system application strange land, system will check that the quantity of current remaining quota is to determine that whether can authorize this combination recovers the authority of data from strange land: if system exists available quota, authorize this request combination quota and available quotas quantity subtracts 1; Otherwise directly return authorization failure, this combination will enter wait state, until other has combined strange land date restoring release quota.

Step 706: after obtaining quota, node will recover data by interzone network from strange land, needs to collect the time delay in data transmission procedure and time-out information at recovery process interior joint.

Under adopting quotas administered mode, the decision flow chart of node system after recovering data success, its concrete steps are as follows:

Step 801: after a node in combination completes recovery process, UDC system needs to check in this area, whether all nodes all complete recovery flow process, recover flow process if completed, and needs to reset regional returning to form.

Step 802: because whole distract completes date restoring, system returns to form regional for reset as " recovery ".

Step 803: system needs the node having checked recovery operation whether to apply for the quota that strange land is recovered from system, if applied for, needs to reclaim quota.

Step 804: system reclaims quota from the node completing recovery, the available quotas number of whole system adds 1.

Further, UDC system can support the quantity dynamically increasing concurrent recovery.After system enables quotas administered, if the time delay of network normally, the concurrent quantity of the recovery of elevator system the time of recovering can be shortened.During each dynamic conditioning, the quota that increase system is supported by system, allows the combination of a wait state to initiate recovery request to strange land.In order to evade the adjustment concussion occurred under boundary condition, system needs the regulation time interval increasing concurrent quantity to amplify.

Further, above-mentioned solution be system automatically according to data packets time delay and time-out perform dynamic conditioning, system also can be supported above-mentioned scheme modifying not to be performed by hand above-mentioned 3 kinds of recovery policies.Keeper can select suitable recovery policy according to the network quality of existing network is manual.The control flow of each strategy is identical with the control mode under automated manner.

The above is only embodiments of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims

1. the control method of a date restoring, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions, it is characterized in that, described control method comprises:

When described Congestion Level SPCC exceeds predetermined congestion control threshold, described control device reduces in described first area simultaneously by the quantity of the node of interzone network recovery data;

Wherein, comprised by the quantity of the node of interzone network recovery data in described minimizing first area simultaneously:

2. control method as claimed in claim 1, is characterized in that, also comprise:

3. control method as claimed in claim 1, is characterized in that,

4. the control method as described in claim 1 or 3, is characterized in that, also comprises:

5. the control device of a date restoring, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions, it is characterized in that, described control device comprises:

First control unit, for when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduces in the first area simultaneously by the quantity of the node of interzone network recovery data;

Wherein, described first control unit comprises:

6. control device as claimed in claim 5, is characterized in that, also comprise:

Second control unit, for when described Congestion Level SPCC does not exceed described congestion control threshold, increases in the first area simultaneously by the quantity of the node of interzone network recovery data.

7. control device as claimed in claim 5, it is characterized in that, described first control unit also comprises:

8. a disaster tolerance system, comprising:

Be arranged on the node in the first area and be arranged on bizonal node;

It is characterized in that, also comprise: the control device as described in any one of claim 5 to 7.

9. the control method of a data Replica, the node being applied to the first area recovers the process of data from bizonal Nodes, wherein, connected by local network between the node of areal, connected by interzone network between the node of different regions, it is characterized in that, described control method comprises:

Described first node sends described performance index to control device, with the described performance index making described control device send based on each node, judge the Congestion Level SPCC of interzone network, and when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduce in described first area simultaneously by the quantity of the node of interzone network recovery data;

Wherein, described first node receives the first control command that described control device issues, described first control command instruction only allows a node in first node combination to pass through interzone network recovery data, described first node combination is the combination of nodes belonging to described first node, and only comprises in the first area for storing the node of identical data in the combination of described first node;

10. control method as claimed in claim 9, is characterized in that,

11. control methods as claimed in claim 9, is characterized in that, also comprise:

12. control methods as described in claim 10 or 11, is characterized in that,

13. 1 kinds of first nodes, described first node is arranged in the first area, it is characterized in that, described first node comprises:

Report unit, for sending described performance index to control device, with the described performance index making described control device send based on each node, judge the Congestion Level SPCC of interzone network, and when described Congestion Level SPCC exceeds predetermined congestion control threshold, reduce in the first area simultaneously by the quantity of the node of interzone network recovery data;

14. first nodes as claimed in claim 13, is characterized in that, also comprise:

15. first nodes as claimed in claim 13, is characterized in that, also comprise:

16. first nodes as described in claims 14 or 15, is characterized in that, also comprise: