CN105827678B - Communication means and node under a kind of framework based on High Availabitity - Google Patents
Communication means and node under a kind of framework based on High Availabitity Download PDFInfo
- Publication number
- CN105827678B CN105827678B CN201510006672.8A CN201510006672A CN105827678B CN 105827678 B CN105827678 B CN 105827678B CN 201510006672 A CN201510006672 A CN 201510006672A CN 105827678 B CN105827678 B CN 105827678B
- Authority
- CN
- China
- Prior art keywords
- node
- cluster
- sub
- data
- data manipulation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Multi Processors (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the present invention provides the communication means and node under a kind of framework based on High Availabitity, company-data lab environment based on High Availabitity, it is only necessary to carry out network communication with a wherein service node for main accessed node, the data manipulation to be carried out can be forwarded to other service nodes in cluster up by the service node, to achieve the purpose that each service node data simultaneously operating.Same operation is being carried out to multiple nodes without main accessed node, reduce repetitive operation, it is established without main accessed node to multiple nodes and is connected to the network and sends identical operation content, on the basis of reduction communication network resource utilization rate, it is few that there is also levels, the simple feature in path, the problem of improving the efficiency of data manipulation, avoid complex management.
Description
Technical field
The present invention relates to the communication means under communication, distributed type assemblies field more particularly to a kind of framework based on High Availabitity
And node.
Background technique
Present distributed type assemblies database generallys use multinode data base concurrency operating mode, such as DB2DPF framework
Deng mass data is distributed to different nodes by certain algorithm to improve the degree of parallelism of read-write operation.Considering to improve
While parallel execution efficiency, in order to guarantee the high availability of data, also requirement can be to other nodes between different nodes
Data are backed up accordingly, in order in the case where there is exception in certain nodes, other nodes can substitute the node after
It is continuous to work, to make the external service of total system that can do as usual progress, this namely so-called distributed type assemblies database
High availability.
In the distributed type assemblies database environment of High Availabitity, due to the data of same content all exist in multiple nodes it is standby
Part, when needing to carry out data manipulation it is necessary to carry out consistency operation to the data on each node, to guarantee each node
Data are consistent, namely under the premise of not losing points cloth parallel work-flow advantage, while guaranteeing the high availability and data of database
Consistency.
Currently, being mainly based on the communication means under High Availabitity framework in order to ensure data consistency, main accessed node is first
Network connection is first passed through to all nodes with the data backup, identical data behaviour then is carried out to each node respectively
Make, to reach the consistent purpose of each node data, network architecture diagram at this time is as shown in Figure 1.
But this communication mode the prior art has at least the following problems:
1, multiple spot is mutually for there are repetitive operations
It, will when the identical data of multiple nodes of the main accessed node into distributed type assemblies database carries out same operation
It is calculated there are identical data and is repeated as many times the case where (number of repetition is depending on number of nodes) carries out, this can greatly improve system money
Source consumption, to seriously reduce the execution efficiency of data manipulation.
2, the one-to-many communication network resource utilization rate of tradition is low
Main accessed node is more to distributed type assemblies data store internal generally by distributed type assemblies database external network
A node, which is established, to be connected to the network and sends identical operation content.Identical data content is repeatedly carried out network transmission, this meeting
Increase network data conveying capacity at several times, network data transmission pressure is considerably increased, to seriously reduce distributed type assemblies
Database external network service efficiency.
3, there are joint behavior Bucket Principles
When main accessed node carries out operating result confirmation to each node, need to wait the operation on multiple nodes all true
Recognize and finish, just can determine that whether this operation is completed.Since the process performance of each node is not quite similar, main accessed node
It must not be not to wait for the nodal operation that performance is worst in multiple nodes to finish, just can confirm that completion this time operates.During this, Qi Taxing
Preferably stagnation wait state can be in by node, distributed type assemblies database entirety service efficiency is low.
Based in another communication means under High Availabitity framework, providing a kind of best land ownership of distributed data duplication
The method and system of weighing apparatus replication overhead and consistency level.By hierarchical fashion gradually by data distribution to subsequent node, from
And determine optimum link.Network architecture diagram at this time is as shown in Fig. 2, in system 10, including multiple is arranged according to predetermined hierarchy
Private server 12, multiple leaf node D-servers 20, multiple intermediate node D servers 16, a root node D-server 14.
It further include multiple resource discovery servers with special domain or weave connection and referred to as O-server 22, distributed O in system 10
Server is across wide area network 18 and D server communication, in addition, other O-servers can also pass through one or more local area networks 24 and D
Server communication.
This under level is distributed by way of realization communicates, although communication performance is obtained relative to the first implementation
Arrived optimization, but there are still levels it is excessive, path is irregular the problems such as, excessively then the efficiency of data manipulation still has level
Bottleneck, path is irregular, and it will cause complex managements.
Summary of the invention
The embodiment of the present invention provides the communication means and node under a kind of framework based on High Availabitity, for improving data manipulation
Efficiency, and reduce the complexity of management.
A kind of communication means under the framework based on High Availabitity, which comprises
Agent node and main accessed node in high-available distributed Cluster Database in randomly selected a subset group
Establish network connection;
The agent node that network connection is established with main accessed node determines what the data manipulation of main accessed node request was related to
Sub-cluster information and the second data operation request information;
With the agent node of main accessed node foundation network connection, in each sub-cluster being related to respectively with data manipulation
Agent node establishes the connection of cluster interior nodes, and sends the second data operation request, wherein the subset that each data manipulation is related to
Agent node in group indicates that node relevant to data manipulation is held in this cluster according to the second data operation request received
Row data manipulation, and at the end of determining the data manipulation of this sub-cluster interior nodes, summarize the data behaviour of this sub-cluster interior nodes
Make implementing result;
The agent node that network connection is established with main accessed node is received for the sub-cluster that each data manipulation is related to
The data manipulation implementing result that agent node in the sub-cluster reports after summarizing, and the data reported to each agent node are grasped
As implementing result summarized after be sent to main accessed node;
Wherein, the sub-cluster is according to preset first rule, to the high-available distributed Cluster Database
In node divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is no less than two
It is a;Agent node in a subset group be according to preset Second Rule, selected from the node of the sub-cluster one
A node.
A kind of node, the node include:
Network connecting module is connected to the network for establishing with main accessed node;
Determining module, for determining sub-cluster information and the second data behaviour that the data manipulation of main accessed node request is related to
Make solicited message, wherein the sub-cluster is according to preset first rule, to the high-available distributed company-data
What the node in library was divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is no less than
Two;
Cluster link block, the generation in each sub-cluster that the data manipulation for determining respectively with determining module is related to
Reason node establishes the connection of cluster interior nodes, and sends the second data operation request that determining module is determined, wherein a subset
Agent node in group is a node selecting from the node of the sub-cluster according to preset Second Rule;
Receiving module, the sub-cluster for being related to for each data manipulation, the agent node received in the sub-cluster converge
The data manipulation implementing result that the General Logistics Department reports, and receive the second data operation request;
Execution module, the second data operation request for being received according to receiving module indicate in this cluster and data
It operates relevant node and executes data manipulation, and at the end of determining the data manipulation of this sub-cluster interior nodes, summarize this subset
The data manipulation implementing result of group's interior nodes is reported to the agent node that network connection is established with main accessed node;
Sending module, the data manipulation implementing result for will receive to receiving module are sent to main visit after summarizing
Ask node.
A kind of node, the node include:
First link block is connected to the network for establishing with main accessed node, and receives main accessed node is sent first
Data operation request;
Processing module, the first data operation request for being received according to the first link block determine main access section
The sub-cluster information and the second data operation request information that the data manipulation of point request is related to;
Second link block, for the generation in a subset group randomly selected in high-available distributed Cluster Database
It manages node and establishes network connection, and sub-cluster information that processing module is determined and the second data operation request information are sent to
With the agent node, wherein the sub-cluster is according to preset first rule, to the high-available distributed cluster number
It is divided, and is divided in obtained sub-cluster according to the node in library, at least one sub-cluster interior joint quantity is many
In two;Agent node in a subset group is selected from the node of the sub-cluster according to preset Second Rule
A node.
The scheme provided according to embodiments of the present invention, the company-data lab environment based on High Availabitity, main accessed node is only
It needs to carry out network communication with a wherein service node, can be turned the data manipulation to be carried out by the service node
Other service nodes in cluster are dealt into up, to achieve the purpose that each service node data simultaneously operating.It is being not necessarily to main visit
It asks that node carries out same operation to multiple nodes, reduces repetitive operation, establish network to multiple nodes without main accessed node and connect
Identical operation content is connect and sends, on the basis of reducing communication network resource utilization rate, it is few that there is also levels, and path is simple
The problem of feature improves the efficiency of data manipulation, avoids complex management.
Detailed description of the invention
Fig. 1 is a kind of network architecture diagram that the prior art provides;
Fig. 2 is a kind of network architecture diagram that the prior art provides;
Fig. 3 is the step flow chart for the communication means that the embodiment of the present invention one provides;
Fig. 4 is the network architecture schematic diagram that the embodiment of the present invention one provides;
Fig. 5 is the step flow chart of communication means provided by Embodiment 2 of the present invention;
Fig. 6 is the node structure schematic diagram that the embodiment of the present invention three provides;
Fig. 7 is the node structure schematic diagram that the embodiment of the present invention four provides.
Specific embodiment
The present invention program provides the grouping chain type forwarding communication plan under a kind of framework based on High Availabitity.The program is based on
The company-data lab environment of High Availabitity.Whole nodes are grouped first inside Cluster Database, main accessed node it is only necessary to
With arbitrarily organized in cluster in certain service node (referred to as agent node) carry out network communication, can be by the node by will be into
The agent node that capable data manipulation is in parallel distributed in other groups is to complete concurrent reading and concurrent writing.Dependent on group node each in cluster
Between proprietary high-speed internal communication network, main accessed node only by communicated with initial agent node can understand in real time it is each
The operation executive condition of group interior nodes.
Below in conjunction with Figure of description, preferred embodiment of the present invention will be described, it should be understood that described herein
Preferred embodiment is only for the purpose of illustrating and explaining the present invention and is not intended to limit the present invention.And in the absence of conflict, this Shen
Please in embodiment and embodiment in feature can be combined with each other.
Embodiment one,
The embodiment of the present invention one provides the communication means under a kind of framework based on High Availabitity, the step process of the communication means
Figure can be as shown in Figure 3, comprising:
Step 101, agent node and main accessed node, which are established, to be connected to the network.
In this step, the agent node in high-available distributed Cluster Database in randomly selected a subset group with
Main accessed node establishes network connection.
It is connected to the network specifically, agent node can be established directly with main accessed node, it can also be by pre-processing node
It establishes and is connected to the network with main accessed node.
Fail if the network connection of agent node and main accessed node is established, mistake can be replied to main accessed node
Information terminates this process to be notified to unusual condition.
Wherein, the sub-cluster is according to preset first rule, to the high-available distributed Cluster Database
In node divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is no less than two
It is a;Agent node in a subset group be according to preset Second Rule, selected from the node of the sub-cluster one
A node.
First preset rules can be, but not limited to as random division, or for according to data in node at least one
The context of specific field is divided.
Second preset rules can be, but not limited to as random selection, or for according to the resource service condition of node,
Resource in a subset group is selected to use least node.
Specifically, when selecting agent node, can be, but not limited to by with lower section in the resource service condition according to node
Formula is realized:
In each node deployment monitoring agent module, be responsible for monitoring the resource service condition of this node, as cpu busy percentage,
IO utilization rate, network usage etc., while periodically from other node isochronous resources service conditions to local preservation, so as to reality
Now according to the resource service condition of node, agent node is selected.
Preferably, the node in a subset group can preset priority, and synchronizing sequence can be preferential according to node
Grade carries out.The resource service condition that each node is saved oneself is synchronous to the upper priority node of oneself, and with most
New time stamp data covers old time stamp data.
Step 102, agent node determine the sub-cluster that data manipulation is related to and the second data operation request.
In this step, the agent node that network connection is established with main accessed node determines the number of main accessed node request
The sub-cluster information and the second data operation request information being related to according to operation.
If agent node is directly established with main accessed node and is connected to the network, in this step, agent node receives master
The first data operation request that accessed node is sent, determines the sub-cluster that data manipulation is related to according to first data operation request
Information and the second data operation request information.
If agent node is established by pretreatment node and main accessed node and is connected to the network, pretreatment node can be used for
The primary data operation requests that main accessed node is sent are received, and primary data operation requests can be carried out to subsequent agency point
Hair, is equivalent to access layer.In this step, agent node can receive the number for the main accessed node request that pretreatment node is sent
According to operation the sub-cluster information and the second data operation request information that are related to, son that the data manipulation of main accessed node request is related to
Cluster information and the second data operation request information are to pre-process node and receive the first data manipulation that main accessed node is sent to ask
It asks, is determined according to first data operation request.
Step 103, agent node establish the connection of cluster interior nodes.
In this step, the agent node of network connection is established with main accessed node, what is be related to respectively with data manipulation is every
Agent node in a sub-cluster establishes the connection of cluster interior nodes, and sends the second data operation request.
Certainly, if the sub-cluster that data manipulation is related to includes establishing the agent node institute of network connection with main accessed node
Sub-cluster, then can be understood as with main accessed node establish network connection agent node and itself establish cluster internal segment
Point connection, and the second data operation request is sent to itself.
Specifically, in this step, the agent node of network connection being established with main accessed node, is related to data manipulation
A subset group in agent node when establishing cluster interior nodes connection failure, indicate the sub-cluster according to preset
Two rules, redefine agent node, and establish cluster interior nodes with the agent node that this is redefined out and connect.
The agent node in sub-cluster that each data manipulation is related to is according to the second data operation request received, instruction
Node execution data manipulation relevant to data manipulation in this cluster, and determining that the data manipulation of this sub-cluster interior nodes terminates
When, summarize the data manipulation implementing result of this sub-cluster interior nodes.
Data manipulation implementing result is sent to main accessed node by step 104, agent node.
In this step, the agent node that network connection is established with main accessed node, is related to for each data manipulation
Sub-cluster receives the data manipulation implementing result reported after the agent node in the sub-cluster summarizes, and will be to each agent node
The data manipulation implementing result reported is sent to main accessed node after being summarized.
Can also include step 103 after step 103 ', step 103 ' execution be not limited to it is as shown in Figure 3 in step
Before rapid 104.
Step 103 ', agent node to main accessed node send data operation request confirmation receive information.
The agent node that network connection is established with main accessed node sends data operation request confirmation to main accessed node and connects
It collects mail and ceases, indicate that main accessed node sends new data operation request.
It should be noted that the process that this step can execute data manipulation with sub-cluster each in step 103 carries out simultaneously.
The schematic network structure that the communication means is related to can be as shown in Figure 4.It is divided into three subsets with Cluster Database
Group, respectively sub-cluster 1, sub-cluster 2 and sub-cluster 3, main accessed node (Client) pass through pretreatment node and random selection
Sub-cluster 1 in agent node (Node1-1) establish for network connection, it is assumed that the data manipulation of main accessed node request relates to
And sub-cluster 2 and sub-cluster 3, Node1-1 can respectively with the agent node Node2-1 in sub-cluster 2, the generation in sub-cluster 3
Reason node Node3-1 establishes the connection of cluster interior nodes.And agent node can indicate section relevant to data manipulation in this cluster
Point executes data manipulation, as shown in figure 4, Node1-1, Node2-1 and Node3-1 can be indicated in this cluster and data manipulation
Relevant node executes data manipulation.
It is illustrated below by the scheme that a specific example provides the embodiment of the present invention one.
Embodiment two,
It is straight with Client with agent node second embodiment of the present invention provides the communication means under a kind of framework based on High Availabitity
It connects for foundation is connected to the network and is illustrated, the step flow chart of the communication means can be as shown in Figure 5, comprising:
Step 201, Client and agent node Node1-1, which are established, to be connected to the network.
In this step, Client is sent in network connecting request, with randomly selected sub-cluster 1 to Cluster Database
Agent node Node1-1 establishes network connection.
Step 202, Client send initial data operation requests.
In this step, Client can send initial data operation requests (the first data operation request) to Node1-1,
Node1-1 determines the sub-cluster that data manipulation is related to according to the request, and can determine to be needed accordingly to other sub-clusters
The data operation request (the second data operation request) of forwarding.
Step 203, Node1-1 establish the connection of cluster interior nodes.
Assuming that Node1-1 determines that sub-cluster that data manipulation is related to is sub-cluster 2 and sub-cluster 3, then in this step,
Node1-1 can be established with the agent node Node2-1 in the sub-cluster 2 and agent node Node3-1 in sub-cluster 3 to be collected
Group's interior nodes high speed connects.
Step 204, Node1-1 determine whether the connection of cluster interior nodes succeeds.
In this step, Node1-1 can according to whether establishing successful connection to determine whether needing in other sub-clusters
The new agent node of interior selection.This step is operation in cluster, transparent to Client.
If Node1-1 determines cluster interior nodes successful connection, second can be sent to the agent node of successful connection
Data operation request.Otherwise, it can indicate that the sub-cluster where the agent node of connection failure reselects agent node, until
Determine cluster interior nodes successful connection.
Step 205, Node1-1 send data operation request confirmation to Client and receive information.
Node1-1 is determining cluster interior nodes successful connection, and sends the second data behaviour to the agent node of successful connection
After requesting, data operation request confirmation can be sent to Client and receives information.Client upon receiving this information, can be with
New data operation request is sent to Cluster Database.
Before Client sends new data operation request, it can be saved with the agency in the sub-cluster that is randomly selected again
Point (may be the agent node except Node1-1) establishes new network connection.
Step 206, each sub-cluster execute data manipulation.
In this step, data manipulation is dispersed in each sub-cluster and executes respectively.Each node can hold data manipulation
Row result is aggregated into the agent node of this sub-cluster.
Step 207, Node1-1 reported information are to Client.
In this step, Node1-1 can receive the data behaviour of this cluster interior nodes reported after each agent node summarizes
Make implementing result, and the data manipulation implementing result of each sub-cluster interior nodes summarized is returned to Client.
The scheme that the embodiment of the present invention one and embodiment two provide is the premise that each nodal parallel executes in cluster
Under, the grouping chain type forwarding communication means under a kind of framework based on High Availabitity is proposed to ensure the consistency of data manipulation.It is logical
It crosses and sub-cluster and agent node is set in the cluster, i.e., agent node is first selected by preset rule, in sub-cluster
Communication and sub-cluster between task transmitting, solve the problems, such as inefficiency present in current communication methods, to make to collect
The performance of each node consistency data manipulation of group performs to optimal.
Based on the same inventive concept with the embodiment of the present invention one, two, node below is provided.
Embodiment three,
The embodiment of the present invention three provides a kind of node, and the structural schematic diagram of the node can be as shown in Figure 6, comprising:
Network connecting module 11 is used to establish with main accessed node and be connected to the network;
Determining module 12 is used to determine the sub-cluster information and the second data that the data manipulation of main accessed node request is related to
Operation requests information, wherein the sub-cluster is according to preset first rule, to the high-available distributed cluster number
It is divided, and is divided in obtained sub-cluster according to the node in library, at least one sub-cluster interior joint quantity is many
In two;
In each sub-cluster that the data manipulation that cluster link block 13 is used to determine with determining module respectively is related to
Agent node establishes the connection of cluster interior nodes, and sends the second data operation request that determining module is determined, wherein a son
Agent node in cluster is a node selecting from the node of the sub-cluster according to preset Second Rule;
Receiving module 14 is used for the sub-cluster being related to for each data manipulation, and the agent node received in the sub-cluster converges
The data manipulation implementing result that the General Logistics Department reports, and receive the second data operation request;
Execution module 15 is used for the second data operation request for receiving according to receiving module, indicates in this cluster and data
It operates relevant node and executes data manipulation, and at the end of determining the data manipulation of this sub-cluster interior nodes, summarize this subset
The data manipulation implementing result of group's interior nodes is reported to the agent node that network connection is established with main accessed node;
Sending module 16 is for being sent to master after summarizing the data manipulation implementing result received to receiving module
Accessed node.
The determining module 12 is specifically used for receiving the first data operation request that main accessed node is sent, according to this first
Data operation request determines the sub-cluster information that data manipulation is related to and the second data operation request information.
The network connecting module 11, which is specifically used for establishing by pretreatment node and main accessed node, to be connected to the network;
The data manipulation that the determining module 12 is specifically used for receiving the main accessed node request that pretreatment node is sent relates to
And sub-cluster information and the second data operation request information, the sub-cluster information that the data manipulation of main accessed node request is related to
It is to pre-process node to receive the first data operation request that main accessed node is sent with the second data operation request information, according to this
What the first data operation request was determined.
The sending module 16, which is also used to send data operation request confirmation to main accessed node, receives information, indicates main visit
Ask that node sends new data operation request.
The agent node that the cluster link block 13 is specifically used in a subset group being related to data manipulation is built
When vertical cluster interior nodes connection failure, indicate that the sub-cluster according to preset Second Rule, redefines agent node, and
Cluster interior nodes are established with the agent node that this is redefined out to connect.
Example IV,
The embodiment of the present invention four provides a kind of node, the structural schematic diagram of the node can with as shown in fig. 7, comprises:
First link block 21 is used to establish with main accessed node and be connected to the network, and receives main accessed node is sent first
Data operation request;
Processing module 22 is used for the first data operation request received according to the first link block, determines main access section
The sub-cluster information and the second data operation request information that the data manipulation of point request is related to;
Second link block 23 be used for in a subset group randomly selected in high-available distributed Cluster Database
Agent node establishes network connection, and the sub-cluster information that processing module is determined and the second data operation request information are sent
Give the agent node, wherein the sub-cluster is according to preset first rule, to the high-available distributed cluster
What the node in database was divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is not
Less than two;Agent node in a subset group is selected from the node of the sub-cluster according to preset Second Rule
The node selected.
Compared with the existing technology, main advantage includes at least the following to the scheme that various embodiments of the present invention provide:
1, Data duplication calculating is avoided, resource is saved
The end Client is without carrying out computing repeatedly work in the present invention program, it is only necessary to which agent node (or pretreatment node) is right
After data carry out one-time calculation, then by agent node the operation requests after calculating are sent to other agent nodes, ensure that
The efficient utilization of the end Client resource.
2, cluster outside network resource utilization rate is high
In the present invention program, Client only needs to send once-through operation request to cluster environment.And in non-conventional mode
It needs operation requests sending n times (n node), cluster outside network resource utilization rate is greatly saved to further increase
Handling capacity of the Client to collection pocket transmission operation requests.
3, the problem of joint behavior Bucket Principle is not present
The present invention program is to be operated by agent node Node1-1 to other agent node Node2-1, Node3-1
The mode of forwarding is requested to carry out, the data manipulation of each node is mutually indepedent, can elect in the case where being abnormal
New agent node continues to execute task, will not send reward data to Client again.Thus constitute each nodal operation
The mutually independent cluster environment of progress.Therefore, even if the performance of different node difference, due to the operation processing of each node
Progress is mutually indepedent, and the good node of performance will not be tied down by the node of performance difference.To ensure that each joint behavior most
The performance of limits.
4, in addition, scheme provided by the invention is also not present, level is excessive, and the efficiency of data manipulation still has bottleneck, with
And the problem of complex management caused by the complexity of path.
It should be understood by those skilled in the art that, embodiments herein can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the application
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the application, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The application is referring to method, the process of equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although the preferred embodiment of the application has been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications can be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the application range.
Obviously, those skilled in the art can carry out various modification and variations without departing from the essence of the application to the application
Mind and range.In this way, if these modifications and variations of the application belong to the range of the claim of this application and its equivalent technologies
Within, then the application is also intended to include these modifications and variations.
Claims (13)
1. the communication means under a kind of framework based on High Availabitity, which is characterized in that the described method includes:
Agent node and main accessed node in high-available distributed Cluster Database in randomly selected a subset group are established
Network connection;
The agent node that network connection is established with main accessed node determines the subset that the data manipulation of main accessed node request is related to
Group's information and the second data operation request information;
The agent node of network connection, the agency in each sub-cluster being related to respectively with data manipulation are established with main accessed node
Node establishes the connection of cluster interior nodes, and sends the second data operation request, wherein in the sub-cluster that each data manipulation is related to
Agent node according to the second data operation request received, indicate node execution number relevant to data manipulation in this cluster
According to operation, and at the end of determining the data manipulation of this sub-cluster interior nodes, the data manipulation for summarizing this sub-cluster interior nodes is held
Row result;
The agent node for establishing network connection with main accessed node receives the son for the sub-cluster that each data manipulation is related to
The data manipulation implementing result that agent node in cluster reports after summarizing, and the data manipulation reported to each agent node is held
Row result is sent to main accessed node after being summarized;
Wherein, the sub-cluster is according to preset first rule, in the high-available distributed Cluster Database
What node was divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is no less than two;One
Agent node in a sub-cluster is according to preset Second Rule, and one selected from the node of the sub-cluster saves
Point.
2. the method as described in claim 1, which is characterized in that establish the agent node of network connection with main accessed node, really
The sub-cluster information and the second data operation request information that the data manipulation of fixed main accessed node request is related to, specifically include:
The agent node that network connection is established with main accessed node receives the first data operation request that main accessed node is sent,
The sub-cluster information and the second data operation request information that data manipulation is related to are determined according to first data operation request.
3. the method as described in claim 1, which is characterized in that randomly selected one in high-available distributed Cluster Database
Agent node and main accessed node in sub-cluster, which are established, to be connected to the network, and is specifically included:
Agent node in high-available distributed Cluster Database in randomly selected a subset group, by pretreatment node with
Main accessed node establishes network connection;
The agent node that network connection is established with main accessed node determines the subset that the data manipulation of main accessed node request is related to
Group's information and the second data operation request information, specifically include:
The agent node of network connection is established with main accessed node, receives the number for the main accessed node request that pretreatment node is sent
According to operation the sub-cluster information and the second data operation request information that are related to, son that the data manipulation of main accessed node request is related to
Cluster information and the second data operation request information are to pre-process node and receive the first data manipulation that main accessed node is sent to ask
It asks, is determined according to first data operation request.
4. the method as described in claim 1, which is characterized in that the agent node of network connection is established with main accessed node, point
The agent node in each sub-cluster not being related to data manipulation is established cluster interior nodes and is connect, and sends the second data manipulation
After request, the method also includes:
The agent node that network connection is established with main accessed node sends data operation request confirmation to main accessed node and receives letter
Breath, indicates that main accessed node sends new data operation request.
5. the method as described in claim 1, which is characterized in that the agent node of network connection is established with main accessed node, point
The agent node in each sub-cluster not being related to data manipulation is established cluster interior nodes and is connect, and specifically includes:
The agent node that network connection is established with main accessed node, agency's section in a subset group being related to data manipulation
When point establishes cluster interior nodes connection failure, indicate that the sub-cluster according to preset Second Rule, redefines agency's section
Point, and establish cluster interior nodes with the agent node that this is redefined out and connect.
6. the method as described in Claims 1 to 5 is any, which is characterized in that first rule is random division, or is root
It is divided according to the context of at least one specific field of data in node.
7. the method as described in Claims 1 to 5 is any, which is characterized in that the Second Rule is random selection, or is root
According to the resource service condition of node, resource in a subset group is selected to use least node.
8. a kind of node, which is characterized in that the node includes:
Network connecting module is connected to the network for establishing with main accessed node;
Determining module, for determining that sub-cluster information that the data manipulation of main accessed node request is related to and the second data manipulation are asked
Seek information, wherein the sub-cluster is according to preset first rule, to the section in high-available distributed Cluster Database
What point was divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is no less than two;
Cluster link block, agency's section in each sub-cluster that the data manipulation for determining respectively with determining module is related to
Point establishes the connection of cluster interior nodes, and sends the second data operation request that determining module is determined, wherein in a subset group
Agent node be a node being selected from the node of the sub-cluster according to preset Second Rule;
Receiving module, the sub-cluster for being related to for each data manipulation receive after the agent node in the sub-cluster summarizes
The data manipulation implementing result reported, and receive the second data operation request;
Execution module, the second data operation request for being received according to receiving module indicate in this cluster and data manipulation
Relevant node executes data manipulation, and at the end of determining the data manipulation of this sub-cluster interior nodes, summarizes in this sub-cluster
The data manipulation implementing result of node is reported to the agent node that network connection is established with main accessed node;
Sending module, the data manipulation implementing result for will receive to receiving module are sent to main access section after summarizing
Point.
9. node as claimed in claim 8, which is characterized in that the determining module is specifically used for receiving main accessed node hair
The first data operation request sent determines the sub-cluster information and second that data manipulation is related to according to first data operation request
Data operation request information.
10. node as claimed in claim 8, which is characterized in that
The network connecting module is specifically used for being connected to the network by pretreatment node and the foundation of main accessed node;
The determining module, the son that the data manipulation specifically for receiving the main accessed node request that pretreatment node is sent is related to
Cluster information and the second data operation request information, the sub-cluster information and second that the data manipulation of main accessed node request is related to
Data operation request information is to pre-process node to receive the first data operation request that main accessed node is sent, according to first number
It is determined according to operation requests.
11. node as claimed in claim 8, which is characterized in that the sending module is also used to send number to main accessed node
Confirm according to operation requests and receive information, indicates that main accessed node sends new data operation request.
12. node as claimed in claim 8, which is characterized in that the cluster link block, be specifically used for data manipulation
When the agent node in a subset group being related to establishes cluster interior nodes connection failure, indicate the sub-cluster according to presetting
Second Rule, redefine agent node, and establish cluster interior nodes with the agent node that this is redefined out and connect.
13. a kind of node, which is characterized in that the node includes:
First link block is connected to the network for establishing with main accessed node, and receives the first data that main accessed node is sent
Operation requests;
Processing module, the first data operation request for being received according to the first link block, determines that main accessed node is asked
The sub-cluster information and the second data operation request information that the data manipulation asked is related to;
Second link block, for being saved with the agency in a subset group randomly selected in high-available distributed Cluster Database
Point establishes network connection, and by sub-cluster information that processing module is determined and the second data operation request information be sent to should
Agent node, the agent node in each sub-cluster which is related to data manipulation respectively establish cluster interior nodes company
It connects, and sends the second data operation request, for the sub-cluster that each data manipulation is related to, receive agency's section in the sub-cluster
The data manipulation implementing result that the point sink General Logistics Department reports, and the data manipulation implementing result reported to each agent node is summarized
After be sent to main accessed node;Wherein, the agent node in sub-cluster that each data manipulation is related to is according to second received
Data operation request indicates that node relevant to data manipulation executes data manipulation in this cluster, and is determining in this sub-cluster
At the end of the data manipulation of node, summarize the data manipulation implementing result of this sub-cluster interior nodes;
Wherein, the sub-cluster is according to preset first rule, in the high-available distributed Cluster Database
What node was divided, and divide in obtained sub-cluster, at least one sub-cluster interior joint quantity is no less than two;One
Agent node in a sub-cluster is according to preset Second Rule, and one selected from the node of the sub-cluster saves
Point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510006672.8A CN105827678B (en) | 2015-01-07 | 2015-01-07 | Communication means and node under a kind of framework based on High Availabitity |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510006672.8A CN105827678B (en) | 2015-01-07 | 2015-01-07 | Communication means and node under a kind of framework based on High Availabitity |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105827678A CN105827678A (en) | 2016-08-03 |
CN105827678B true CN105827678B (en) | 2019-03-05 |
Family
ID=56514740
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510006672.8A Active CN105827678B (en) | 2015-01-07 | 2015-01-07 | Communication means and node under a kind of framework based on High Availabitity |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105827678B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107295080B (en) * | 2017-06-19 | 2020-12-18 | 北京百度网讯科技有限公司 | Data storage method applied to distributed server cluster and server |
CN109714183A (en) * | 2017-10-26 | 2019-05-03 | 阿里巴巴集团控股有限公司 | Data processing method and device in a kind of cluster |
CN109189575A (en) * | 2018-08-20 | 2019-01-11 | 北京奇虎科技有限公司 | A kind of Explore of Unified Management Ideas and device of more OpenStack clusters |
CN110896404B (en) * | 2018-09-12 | 2021-09-14 | 华为技术有限公司 | Data processing method and device and computing node |
CN112398905B (en) * | 2020-09-28 | 2022-05-31 | 联想(北京)有限公司 | Node and information synchronization method |
CN112395334A (en) * | 2020-11-30 | 2021-02-23 | 浪潮云信息技术股份公司 | Method for dividing centerless distributed database cluster into a plurality of logic sub-clusters |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101667034A (en) * | 2009-09-21 | 2010-03-10 | 北京航空航天大学 | Scalable monitoring system supporting hybrid clusters |
CN102355413A (en) * | 2011-08-26 | 2012-02-15 | 北京邮电大学 | Method and system for unifying message space on large scale in real time |
CN102739799A (en) * | 2012-07-04 | 2012-10-17 | 合一网络技术(北京)有限公司 | Distributed communication method in distributed application |
CN103617231A (en) * | 2013-11-26 | 2014-03-05 | 国家电网公司 | Large data management system |
CN103647797A (en) * | 2013-11-15 | 2014-03-19 | 北京邮电大学 | Distributed file system and data access method thereof |
CN103647820A (en) * | 2013-12-09 | 2014-03-19 | 华为数字技术(苏州)有限公司 | Arbitration method and arbitration apparatus for distributed cluster systems |
CN103905544A (en) * | 2014-03-25 | 2014-07-02 | 华为技术有限公司 | Cluster storage system and distribution method of nodes in cluster storage system |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9497281B2 (en) * | 2013-04-06 | 2016-11-15 | Citrix Systems, Inc. | Systems and methods to cache packet steering decisions for a cluster of load balancers |
-
2015
- 2015-01-07 CN CN201510006672.8A patent/CN105827678B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101667034A (en) * | 2009-09-21 | 2010-03-10 | 北京航空航天大学 | Scalable monitoring system supporting hybrid clusters |
CN102355413A (en) * | 2011-08-26 | 2012-02-15 | 北京邮电大学 | Method and system for unifying message space on large scale in real time |
CN102739799A (en) * | 2012-07-04 | 2012-10-17 | 合一网络技术(北京)有限公司 | Distributed communication method in distributed application |
CN103647797A (en) * | 2013-11-15 | 2014-03-19 | 北京邮电大学 | Distributed file system and data access method thereof |
CN103617231A (en) * | 2013-11-26 | 2014-03-05 | 国家电网公司 | Large data management system |
CN103647820A (en) * | 2013-12-09 | 2014-03-19 | 华为数字技术(苏州)有限公司 | Arbitration method and arbitration apparatus for distributed cluster systems |
CN103905544A (en) * | 2014-03-25 | 2014-07-02 | 华为技术有限公司 | Cluster storage system and distribution method of nodes in cluster storage system |
Also Published As
Publication number | Publication date |
---|---|
CN105827678A (en) | 2016-08-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105827678B (en) | Communication means and node under a kind of framework based on High Availabitity | |
US9489443B1 (en) | Scheduling of splits and moves of database partitions | |
CN105933408B (en) | A kind of implementation method and device of Redis universal middleware | |
CN111327681A (en) | Cloud computing data platform construction method based on Kubernetes | |
CN109656911A (en) | Distributed variable-frequencypump Database Systems and its data processing method | |
CN105677836A (en) | Big data processing and solving system simultaneously supporting offline data and real-time online data | |
CN109933631A (en) | Distributed parallel database system and data processing method based on Infiniband network | |
CN107450855B (en) | Model-variable data distribution method and system for distributed storage | |
CN110489225A (en) | A kind of service expansion method, device and equipment based on message queue | |
EP3442201B1 (en) | Cloud platform construction method and cloud platform | |
US10761869B2 (en) | Cloud platform construction method and cloud platform storing image files in storage backend cluster according to image file type | |
CN106161524A (en) | A kind of resource downloading method based on distributed resource node and device | |
CN113687964B (en) | Data processing method, device, electronic equipment, storage medium and program product | |
CN107645410A (en) | A kind of virtual machine management system and method based on OpenStack cloud platforms | |
CN103581332A (en) | HDFS framework and pressure decomposition method for NameNodes in HDFS framework | |
CN114048186A (en) | Data migration method and system based on mass data | |
CN107528871A (en) | Data analysis in storage system | |
Fazul et al. | Improving data availability in HDFS through replica balancing | |
EP3349416B1 (en) | Relationship chain processing method and system, and storage medium | |
Liao et al. | Efficient time-evolving stream processing at scale | |
CN116226067A (en) | Log management method, log management device, processor and log platform | |
Yongdnog et al. | A scalable and integrated cloud monitoring framework based on distributed storage | |
Liu et al. | Edge node data replica management method for distribution Internet of Things | |
CN106557492A (en) | A kind of method of data synchronization and device | |
CN109558214A (en) | Host method for managing resource, device and storage medium under isomerous environment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |