CN103581276B - Cluster management device, system, service customer end and correlation method - Google Patents
Cluster management device, system, service customer end and correlation method Download PDFInfo
- Publication number
- CN103581276B CN103581276B CN201210281817.1A CN201210281817A CN103581276B CN 103581276 B CN103581276 B CN 103581276B CN 201210281817 A CN201210281817 A CN 201210281817A CN 103581276 B CN103581276 B CN 103581276B
- Authority
- CN
- China
- Prior art keywords
- network access
- service
- access address
- service server
- management device
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Abstract
The invention discloses a kind of cluster management device, system, service customer end and correlation method.The cluster management device includes:Receive and memory module, for receiving and store from the respective network access address of each service server sent of each service server in the cluster of management;Sending module, for sending at least one network access address that reception and memory module are stored to service customer end.Further, the cluster management device also includes:Monitor module, the running state information for monitoring each service server;Judge update module, for judging whether the service server corresponding to the network access address that reception and memory module are stored can provide service to service customer end based on running state information, stored network access address is updated according to judged result.The cluster management device can be avoided as much as concentrating the brought network traffics bottleneck of forwarding by the cluster management such as LVS servers relevant apparatus because of access request.
Description
Technical field
The present invention relates to network technology information, more particularly to a kind of cluster management device, system and method and, business client
End and correlation method.
Background technology
With continuing to develop for internet scale, the network user and various network applications are increased rapidly so that network service
The load capacity of device turns into bottleneck and weak link in service gradually.High capacity is undertaken according to single server, then very
The single server of very high cost may be needed could meet required high-load performance.Even, at all cannot be from market
It is upper to purchase to the server that disclosure satisfy that required high-load performance.
Therefore, some are recently proposed using various relatively inexpensive and stable performance high availability cluster system.In order to
Ensure the reliability of the high availability of group system, expandability, manageability, and cluster management system, it is proposed that some
Cluster organization to management etc. related scheme.Below with by Linux virtual server(LVS servers, Linux Virtual
Server)As a example by the group system of structure(With reference to Fig. 1)To illustrate.
LVS servers are typically deployed as load-balanced server(director server), the load with LVS servers
Balanced device constitutes cluster management device 20.After cluster management device 20 receives the request that client 10 is sent, please by what is received
Ask and be transmitted to real server 30 (Real Server).Then, real server 30(Also known as service server)Process and ring
The request that Ying Yu is received, will return to cluster management device 20 as the response results of the request.Finally, cluster management device
Response results are given to client 10 by 20.So, client is obtained by the distribution and forwarding capability of cluster management device 20
Response results from real server 30.
It can be seen that, because client 10 is required to be forwarded just by cluster management device 20 to the request of some collection pocket transmission
Real server 30 can be reached, and real server 30 also will be by cluster management device to the response results that client 10 is returned
20 can just arrive at client 10, and this causes that the flow by cluster management device 20 is huge, and then likely results in cluster management
Device 20 turns into the bottleneck that whole group flow comes in and goes out.
Additionally, using the cluster management device 20 of LVS servers to be mainly responsible for being forwarded flow to reach shunting
Purpose, different services are not distinguished.Therefore, if multiple clusters are managed, then multiple LVS servers are needed to come for not
Same cluster provides management and forwards, and results in the need for deployment multiple individually cluster management devices 20.
The content of the invention
One of technical problems to be solved by the invention be need provide one kind can be avoided as much as because access please
Ask and concentrate the cluster management device of the brought network traffics bottleneck of forwarding by the cluster management such as LVS servers relevant apparatus, be
System, service customer end and correlation method.
In order to solve the above-mentioned technical problem, the invention provides a kind of cluster management device.The cluster management device includes:
Receive and memory module, for receiving and store from each described business service sent of each service server in the cluster of management
The respective network access address of device;Sending module, for sending the reception and memory module storage extremely to service customer end
A few network access address.
Further, the cluster management device also includes:Monitor module, the operation shape for monitoring each service server
State information;Judge update module, the network for judging the reception and memory module storage based on the running state information
Whether the service server corresponding to reference address can provide service to service customer end, updated according to judged result and deposited
The network access address of storage;Wherein, at least one network access address that the sending module sends to service customer end
It is to judge that update module is judged as to be provided for service customer end the network access address of service by described.
Further, the sending module is further used for according to the service customer end for obtaining service server
The request or active of network access address send the network access address of the storage to the service customer end.
Further, the cluster management device is further used for the network access ground for receiving the reception and memory module
Location is managed and is stored with virtual tree construction.
Further, when the cluster management device has managed multiple service server clusters, it is further used for difference
The network access address of the service server in service server cluster stores the different subtrees in the virtual tree construction
In.
Further, it is described monitor the running state information monitored of module include on service server self performance and/
Or the information of loading condition.
According to another aspect of the present invention, the service customer end in a kind of cluster is additionally provided.The service customer end includes:
Second sending module, for sending the network access ground for obtaining service server to the cluster management device in the cluster
The request of location;Second receives and memory module, for receiving and stores the cluster management device according to the request or master
The network access address of the dynamic service server sent;Address determination module, for the network access address that will be stored
One of be defined as objective network reference address;Connection establishment module, for business corresponding with the objective network reference address
Server sets up connection.
Further, the service customer end, also includes:Judge module, for being managed accessing the cluster management device
During the cluster of reason, judge whether the service customer end locally stores at least one network access ground of the service server
Location, if being judged as NO, runs the second sending module, conversely, then running connection establishment module.
Further, the address determination module, is further used in business service corresponding with objective network reference address
When device sets up network connection failure, objective network reference address is redefined.
Further, each time of the service server cluster for being managed the cluster management device in the service customer end
In access, each network access address that will be stored by the way of poll or using the order determined based on client terminal attribute
It is defined as the objective network reference address successively.
Further, it is described second receive and memory module be further used for from the cluster management device receive on
The corresponding service server self performance of network access address list and/or the relevant information of loading condition;And the connection
Set up module and be based further on service server self performance corresponding with the network access address list and/or loading condition
One of network access address that relevant information will be stored is defined as the objective network reference address.
According to another aspect of the present invention, a kind of cluster management system is additionally provided.The cluster management system includes basis
Cluster management device described in service customer end and above-mentioned technical proposal described in above-mentioned technical proposal and its collection for being managed
Each service server in group.
According to another aspect of the present invention, a kind of cluster management method is additionally provided.The cluster management method includes:Cluster
Managing device receives and stores the respective net of each described service server sent from each service server in the cluster of management
Network reference address;Service customer end sends the network access address for obtaining service server to the cluster management device
Request;The cluster management device according to it is described acquisition service server network access address request to the business client
End sends at least one network access address of its own storage;The service customer end is received and stores the cluster management dress
The network access address of the service server sent is put, one of its own network access address for storing is defined as objective network
Reference address, service server corresponding with the objective network reference address sets up connection.
Further, the cluster management method also includes:The cluster management device monitors the operation shape of each service server
State information, judges whether the service server corresponding to the network access address of storage can be to based on the running state information
Service customer end provides and services and update stored network access address according to judged result;The cluster management device exists
After updating stored network access address, being judged as to service customer end transmission can be for service customer end provides service
Network access address.
Further, the running state information is also included on service server corresponding with network access address list certainly
Body performance and/or the relevant information of loading condition;And
The service customer end be based on service server self performance corresponding with the network access address list and/or
One of network access address that the relevant information of loading condition will be stored is defined as the objective network reference address.
Further, the network access address of reception is managed and is deposited by the cluster management device with virtual tree construction
Storage, when the cluster management device has managed multiple service server clusters, by the business in different business server cluster
The network access address of server is stored in the different subtrees in the virtual tree construction.
According to another aspect of the present invention, a kind of management method of cluster management device is additionally provided.Cluster management device
The management method include:Receive and store and sent from each service server in the cluster managed by the cluster management device
The respective network access address of each described service server;At least one stored network access is sent to service customer end
Address.
Further, the management method of the cluster management device also includes:Monitor the running status of each service server
Information;Judge whether service server corresponding to stored network access address can be to based on the running state information
Service customer end provides and services and update stored network access address according to judged result;Wherein, it is described to business visitor
The step of family end sends at least one stored network access address specifically includes:Sent to service customer end and be judged as energy
The network access address of service is enough provided for service customer end.
According to another aspect of the present invention, a kind of access method of the service customer end in cluster is additionally provided.The business
The access method of client includes:The network sent for obtaining service server to the cluster management device in the cluster is visited
Ask the request of address;Receive and store the network access address of the service server that the cluster management device is sent;To be deposited
One of described network access address of storage is defined as objective network reference address;According to the objective network reference address with it is described
The corresponding service server of objective network reference address sets up connection.
Compared with prior art, one or more embodiments of the invention can have the following advantages that:Due to client 10
The relevant information of the running status of each service server 30 in the management cluster is being obtained in that from cluster management device 40
In the case of, can directly be set up with the service server 30 to be accessed and be connected, all access requests can be effectively prevented from
During the flow shunt for being brought is connected to the multiple that each client 10 is set up with each service server 30 respectively, so as to the greatest extent
Possibly avoid because access request concentrates the brought network flow measuring bottle of forwarding by the cluster management such as LVS servers relevant apparatus
Neck.
Other features and advantages of the present invention will be illustrated in the following description, also, the partly change from specification
Obtain it is clear that or being understood by implementing the present invention.The purpose of the present invention and other advantages can be by specification, rights
Specifically noted structure is realized and obtained in claim and accompanying drawing.
Brief description of the drawings
Accompanying drawing is used for providing a further understanding of the present invention, and constitutes a part for specification, with reality of the invention
Apply example to be provided commonly for explaining the present invention, be not construed as limiting the invention.In the accompanying drawings:
Fig. 1 is a kind of structural representation of the cluster management system according to prior art;
Fig. 2 is the structural representation of cluster management system according to a first embodiment of the present invention;
The cluster management device that Fig. 3 shows according to a first embodiment of the present invention carries out the flow chart of cluster management;
Fig. 4 shows the structural representation of cluster management device according to a first embodiment of the present invention;
The client that Fig. 5 shows according to a first embodiment of the present invention carries out the flow chart of Operational Visit;
Fig. 6 shows the structural representation of client according to a first embodiment of the present invention;
Fig. 7 shows the structural representation of cluster management system according to a second embodiment of the present invention;
Fig. 8 shows the structural representation of cluster management system according to a third embodiment of the present invention.
Specific embodiment
Describe embodiments of the present invention in detail below with reference to drawings and Examples, how the present invention is applied whereby
Technological means solves technical problem, and reaches the implementation process of technique effect and can fully understand and implement according to this.Need explanation
As long as not constituting conflict, each embodiment in the present invention and each feature in each embodiment can be combined with each other,
The technical scheme for being formed is within protection scope of the present invention.
In addition, can be in the such as one group department of computer science of computer executable instructions the step of the flow of accompanying drawing is illustrated
Performed in system, and, although logical order is shown in flow charts, but in some cases, can be with different from herein
Order perform shown or described step.
First embodiment
The present embodiment is illustrated below with reference to Fig. 2 to Fig. 6.
Fig. 2 shows the schematic diagram of cluster management system according to an embodiment of the invention.
Cluster management device 40 and the network connection of each service server 30, each industry that reception is sent from each service server 30
The network access address of business server 30.For example can be using the network access address of each service server 30 storage as network access
It is address list, default for representing cluster as one in the tree-shaped virtual file of Zookeeper (being described in detail below)
Each back end under the virtual root node of the cluster that managing device 40 is managed etc..Storage mode can be selected flexibly.Business takes
The be engaged in network access address of device 30 may include combination of domain names, IP address or both and port numbers etc..Each business clothes
Business device 30 is located in the cluster that cluster management device 40 is managed, and is responsible for providing network service to client 10.
Cluster management device 40 is also with the network connection of client 10 with to the transmission network reference address list of client 10.This
Sample, client 10, can be based on from cluster management device after network access address list is obtained from cluster management device 40
40 network access address list, in the case where being forwarded without cluster management device 40 with the service server 30 to be accessed
Network connection directly is set up, service request is sent with to one of each service server 30, so as to obtain in the network to be accessed
Hold.
Describe the cluster management device 40 according to the present embodiment and its treatment in detail with reference to Fig. 3 and Fig. 4 further below.
Cluster management device 40 as shown in Figure 4 includes reception and memory module 41, sending module 42, monitors module 43, judges to update mould
Block 44.It should be noted that the annexation being not limited to shown in Fig. 4, can be connected with each other between each module.
The following detailed description of each step that cluster management is carried out according to the cluster management device 40 of the present embodiment.
Step S310, is received and memory module 41 is received and stored from each industry in the cluster that cluster management device is managed
The respective network access address of each service server 30 that business server 30 is sent.For example, each service server 30 is to cluster pipe
Reason device 40 is registered(Send)The IP ground for providing network service to service customer end 10 (abbreviation client 10) of its own
Location and port numbers (port), cluster management device 40 receive and store the IP (Internet that each service server 30 is sent
Protocol) address and port numbers.These network access addresses come can be received to store as network access address list.More
Preferably, can be cooperateed with using Zookeeper technologies, the IP address of each business service port 30 and port numbers can respectively be deposited
Storage specifically, each network access address is divided as a back end (znode) in the tree-shaped virtual file of Zookeeper
Do not store into the tree-shaped virtual file of Zookeeper one it is default for representing the cluster that cluster management device 40 is managed
Virtual root node under each back end in, be designated as "/A " (referring to Fig. 7).Because Zookeeper allows multiple to be distributed in not
Cooperated based on shared, similar standard file system a tree-shaped Virtual File System with the process on server.
Each back end in Virtual File System is referred to a znode, and each znode can be data correlation to itself
Or its child node, therefore, when being cooperateed with using Zookeeper technologies, it is possible to reduce the program development of technical staff
Amount, preferably avoids data access conflict, can preferably solve Data Concurrent modification, the data being likely to occur under distributed environment
The problem of Backup and Restore and data consistency.
However, for purposes of illustration only, mainly being illustrated by taking network access address list as an example below.
Step S320, sending module 42 sends each business service by receiving and memory module 41 is stored to client 10
The network access address of device 30.Sending module 42 can push each business of stored at least one and take to each client 10 on one's own initiative
The network access address of business device 30, it is preferable that sending module 42 can be according to the net for obtaining service server of client 10
The request of network reference address to 10 sends the network access address of each service server 30 stored up.
Step S330, alternatively, monitors the running state information of the monitoring service server 30 of module 43.Specifically, can supervise
Listen with receive and network access address that memory module 41 is stored corresponding to service server 30 running state information.Fortune
Whether row status information is included on the available information of service server.Usually, if listening to certain service server 30
Running state information, then can determine whether as the service server 30 is available, that is, service can be provided to service customer end 10.Example
Such as, cluster management device 40 periodically can receive the information on running status from service server 30(Below the also abbreviation heart
Hop-information).By monitoring the snoop-operations of module 43, each business clothes in service server cluster can be in real time known as far as possible
The change of the network access address of business device.
Additionally, the running state information that reception and memory module 43 are monitored may also include on service server certainly
The information of body performance and/or loading condition, so, client 10 can from cluster management device 40 obtain these performances and/or
The information of loading condition, and then select to be accessed 30 based on the information of these performances and/or loading condition.So, can be with
Load balance is better achieved.After step S330, into step S340.
Step S340, judges that update module 44 judges that the reception and memory module 41 are deposited based on the running state information
Whether the service server corresponding to the network access address of storage can provide service and according to judgement knot to the service customer end
Fruit updates stored network access address.If more specifically, monitoring less than a certain network access in each network access address
During the information that the corresponding service server 30 in address is sent, then it is judged as that the service server 30 is unavailable.Then, judge to update
Module 44 can update stored network access address according to judged result.The place of the network access address that the renewal is stored
Managing to be:The network access that network access address corresponding with disabled service server 30 is stored from step S310
In address list delete, or can in network access address list with the corresponding network access of disabled service server 30
Address is marked.Additionally, also can be by monitoring the machine of delaying of each service server 30, overburden, fault warning, restarting alarm etc.
Running status come obtain each service server 30 respectively whether available information.Subsequently into step S350.
Step S350, sending module 42 to service customer end 10 send by it is described judge update module be judged as can be should
Service customer end provides the network access address of service.If monitoring the heartbeat message less than a certain service server 30, can be
In S340 by the network access address of the service server 30 delete or be marked as it is impossible, so as to client 10
When sending network access address, the address of the service server 30 is not retransmited, but only send and be judged update module 44 and sentence
Break as the network access address of service can be provided for service customer end.The network that so each client 10 can be stored to it is visited
Ask that address list is updated.It should be noted that sending module 42 can use asynchronous notifications mechanism, make service customer end saturating
Bright use service server.More specifically, after judging that update module 44 is updated operation, sending module 42 can be immediately
The network access address for being stored after updating is sent to client 10, but it is also possible to which a certain moment afterwards retransmits, because
Even if not send immediately, client 10 is due to having saved network access address list therefore still can normally access industry
Business.
It should be noted that the order between above step S320 and step S330 and S340 can be arbitrary.
Fig. 6 shows the structural representation of the client 10 according to the present embodiment.Including judge module 11, the second sending module
12nd, second receive and memory module 13, address determination module 14 and connection establishment module 15.It should be noted that being not limited to Fig. 6
The annexation for showing, can be connected with each other between each module.
Describe the client 10 and its access cluster management device according to the present embodiment in detail with reference to Fig. 5 and Fig. 6
40 managed cluster when each step.
Step S510, judge module 11 judges client 10 in the cluster that cluster management device to be accessed 40 is managed
Whether network access address list is locally stored, the network access address list includes the cluster that cluster management device 40 is managed
Each service server 30 network access address, if locally not storing the network access address list, into step S520;
Conversely, into step S540.
Step S520, the second sending module 12 sends the net for the service server 30 for obtaining to cluster management device 40
The request of network reference address.
Step S530, the second reception and memory module 13 receive and store the network visit that the cluster management device 40 is sent
Ask address.More specifically, receive and request of the storage cluster managing device 40 to service server 30 respond and to client
The network access address list that end 10 is sent.
One of step S540, network access address that address determination module 14 will be stored are defined as the business to be accessed
The objective network reference address of server 30.
Step S550, connection establishment module 15 based on determined by objective network reference address to identified target network
The corresponding service server 30 of network reference address sends service request to obtain the Web content to be accessed.
More specifically, when client 10 will access the cluster that cluster management device 40 is managed, by each step S530
One of network access address for being stored is defined as objective network reference address, business clothes corresponding with objective network reference address
Business device 30 sets up network connection, is then sent to service request to obtain the Web content to be accessed.So, can be in nothing
In the case of need to being forwarded by cluster management device 40, directly using the network access address access service clothes of service server 30
Business device 30.
It should be noted that S510 is optional, using step S510, client 10 is without per secondary to cluster management
The cluster that device 40 is managed obtains network access address list from cluster management device 40 when sending service request, but
The network access address list stored in local memory can be utilized to back up, can so reduce network overhead and delay.
Preferably, in being accessed at each time of the cluster that client 10 is managed cluster management device 40, using poll
Mode uses each network access address that the order determined based on client terminal attribute will be stored successively as objective network
Reference address.
When using polling mode, certain that client 10 will not be in the cluster that is managed of central access cluster management device 40
One service server 30, therefore can preferably realize load balance.Preferably due to when by the way of poll, client
End 10 always since the First service server 30 corresponding to the network access address for being stored access, therefore, it can into
One step is conducted interviews using the order determined based on the attribute of client 10, for example, determine client 10 based on the attribute of client 10
The First service server 30 of wanted poll, because the attribute of each client 10 is different, can so cause wanted poll
First service server 30 is different, so that load is further balanced.
Additionally, in step S530, the second reception and memory module 13 can also be received and net from cluster management device 40
The information relevant with performance and/or loading condition of the corresponding service server 30 of network reference address list.So, can be based on
Service server self performance corresponding with the network access address list and/or the relevant information of loading condition are by each step
One of the network access address stored in rapid S530 corresponding service server 30 is used as objective network reference address.
Additionally, in step S530, the second reception and memory module 13 are not in the reception cluster management device blocked
40 information for sending, but asynchronous carrying out receives, and asynchronous the conditions associated of service server is updated so that
10 can obtain network access address with indiscriminate use S540.
Additionally, address determination module 14 sets up network connection in service server 30 corresponding with objective network reference address
During failure, objective network reference address is redefined.If all accessing failure, indicate that all service servers are current neither
Can use.
Additionally, client 10 can use double buffering(buffer)Storage, each buffer is used for from cluster management
The network access address that device 40 is received, so can effectively reduce the network access address that exclusive reference stored needs frequency
The performance loss of numerous locking.
Client 10 can be using the network access address stored in default order successively each step S530.
It is that client 10 provides each of service to be described in detail with reference to Fig. 6 according to the service server 30 of the present embodiment
Step.
Step S610, the network access address of itself is sent to default cluster management device 40.For example, by the IP of itself
Address and port numbers are registered in cluster management device 40 fake root for representing the cluster that cluster management device 40 is managed
(Fig. 4 is referred under node).
Step S620, the heartbeat message on itself is sent to cluster management device 40, as long as there have heartbeat to indicate that to be available.
For example, can be sent on itself whether available message to cluster management device 40 periodically or according to predetermined interval.This
Outward, the information on self performance and/or loading condition can also be sent to cluster management device 40.
Step S630, the service request at customer in response end 10 sends the content that client 10 is asked with to client 10.
To sum up, because the volume of transmitted data between client 10 or service server 30 and cluster management device 40 does not include visitor
The business tine data that family end 10 is asked to service server 30, therefore data volume greatly reduces, so as to avoid because of cluster pipe
Data bottlenecks phenomenon caused by the data volume of reason device 40 is excessive.
Additionally, when the cluster service scarce capacity of service server 30, can be in the situation for not influenceing current system to run
Under, directly(For example by modes such as hot plugs)Increase service server 30, the service server 30 for newly increasing is to cluster management
Device 40 sends its network access address, and now cluster management device 40 just can at once monitor this event, and by its network
On reference address increase/A nodes, then notify that client 10 updates stored network access address list, so, client
10 soon can be assigned to newly-increased service server 30 a part of service request.So, cluster can quickly be eliminated
Overload problem.
Second embodiment
With reference to Fig. 7, the cluster management system according to the present embodiment includes multiple according to the cluster management of first embodiment dress
40 are put, a cluster for cluster management server is constituted, by the cluster of cluster management server as each service server 30
Manager.Preferably, the cluster of cluster management server includes odd number platform(Such as 3 or 5)Cluster management device 40.When one
When the network access address list of more than half cluster management device 40 is identical, this cluster management device of more than half can be utilized
40 network access address list updates remaining cluster management device 40.The fault-tolerance of whole system can be so improved,
If sub-fraction cluster management device malfunctions, other machines remains to complete the management role to service server 30.
3rd embodiment
With reference to Fig. 8, the cluster management system according to the present embodiment can utilize a cluster pipe according to first embodiment
Reason device 40 is simultaneously multiple different business using the cluster constituted according to a cluster management device 40 of second embodiment
Server cluster provides management service.
Service server according to the present embodiment is substantially the same with above-described embodiment, will not be repeated here.Additionally, client
Most of step with cluster management device 40 is also substantially the same with above-described embodiment.Hereafter only illustrate different from previous embodiment
Part.
When cluster management device 40(Or the cluster of cluster management device 40)Multiple different business services can simultaneously be managed
Device cluster, now, storage is distinguished by the network access address of the service server of different business server cluster.For example, storage
It is different lists, array or the different subtrees being stored as under same virtual tree, to show differentiation.
The network of the service server for obtaining that client sends in step S520 to cluster management device 40 is visited
Asking the request of address includes and the cluster to be accessed or the business to be accessed(Service)Information.
Then, in step S320, cluster management device 40 be based on that the request includes with the cluster to be accessed or to visit
The business asked(Service)Response of the information to be determined as the request to send network access address to client.
With reference to Fig. 8, realized system that employs Zookeeper.Wherein, service server A1 and service server A2 to
Each customer end B 2 provides a kind of business, and service server A3 provides another business to customer end B 1.
Due to when B2 sends network access Address requests, being included in request on wanting access service server A 1 and industry
The information of the business that business server A 2 is provided, therefore the cluster that cluster management device 40 is constituted returns to virtual root node to it
The network access address of each node under "/C "(IP address and port numbers);And when B1 sends network access Address requests, please
Include the information on the business for wanting access service server A 3 to be provided in asking, therefore cluster management device 40 is constituted
Cluster returns to the network access address of each node under virtual root node "/A " to it(IP address and port numbers).Such client
B1 and B2 can obtain correct network access address, therefore, it is possible to access the service server of each of which needs.
In the system for management business server cluster for utilizing the Clustering mechanism of LVS to set up, shunted just for
Same server and interface, the cluster if necessary to provide different services for multiple provide management service, then need using many
Individual LVS servers or LVS server clusters.But the cluster management device 40 or cluster management device collection provided using the present embodiment
Group, if the cluster that multiple provides different services provides service, only needs cluster management device 40 by the network of different clusters
Reference address is stored respectively, for example, the different znode nodes in being registered to cluster management device 40, as shown in Figure 7.So, A1
With A2 for B2 provides service, A3 provides service for B1, is not in interference between different services.Also, no matter how many is individual
Service cluster provides service and how many client using service, can be filled with same cluster management device 40 or cluster management
The cluster (cluster that 3 cluster management devices 40 as shown in Figure 7 are constituted) for putting composition is managed.
In brief, cluster management device 40 can carry out coordinated management using Zookeeper technologies, using Zookeeper
The network access address that the reception and memory module are received is managed and is stored with virtual tree construction.In cluster management
When device 40 has managed multiple service server clusters, the network of the service server in different business server cluster can be visited
In asking different subtrees of the address storage in virtual tree construction.
Those skilled in the art should be understood that above-mentioned of the invention each module or each step can use general calculating
Device realizes that they can be concentrated on single computing device, or be distributed in the network that multiple computing devices are constituted
On, alternatively, they can be realized with the executable program code of computing device, it is thus possible to be stored in storage
Performed by computing device in device, or they are fabricated to each integrated circuit modules respectively, or will be many in them
Individual module or step are fabricated to single integrated circuit module to realize.So, the present invention be not restricted to any specific hardware and
Software is combined.
Although disclosed herein implementation method as above, described content is only to facilitate understanding the present invention and adopting
Implementation method, is not limited to the present invention.Any those skilled in the art to which this invention pertains, are not departing from this
On the premise of the disclosed spirit and scope of invention, any modification and change can be made in the formal and details implemented,
But scope of patent protection of the invention, must be still defined by the scope of which is defined in the appended claims.
Claims (13)
1. a kind of cluster management device, it is characterised in that including:
Receive and memory module, for receiving and store from each described business sent of each service server in the cluster of management
The respective network access address of server;
Sending module, for sending at least one network access ground that the reception and memory module are stored to service customer end
Location,
Also include:Monitor module, the running state information for monitoring each service server;
Judge update module, the network access for judging the reception and memory module storage based on the running state information
Whether the service server corresponding to address can provide service to service customer end, update what is stored according to judged result
Network access address;
Wherein, the sending module at least one network access address that service customer end sends is judged more by described
New module is judged as to be provided for service customer end the network access address of service,
Wherein, based on the network access address list from cluster management device, in situation about being forwarded without cluster management device
Under, service customer end directly sets up network connection with the service server to be accessed, and is sent with to one of each service server
Service request, so that the Web content to be accessed is obtained,
The cluster management device is further used for the network access address for receiving the reception and memory module with virtual
Tree construction is managed and stores, and when the cluster management device has managed multiple service server clusters, is further used for
The network access address of the service server in different business server cluster is stored in the virtual tree construction not
In with subtree, and coordinated management is carried out to the data in virtual tree construction using Zookeeper technologies.
2. cluster management device according to claim 1, it is characterised in that the sending module is further used for according to institute
State the request of the network access address for obtaining service server of service customer end or actively to the service customer end
Send the network access address of the storage.
3. cluster management device according to any one of claim 1 to 2, it is characterised in that the monitoring module is supervised
The running state information listened includes the information on service server self performance and/or loading condition.
4. the service customer end in a kind of cluster, it is characterised in that including:
Second sending module, the network for being sent to the cluster management device in the cluster for obtaining service server is visited
Ask the request of address;
Second receive and memory module, for receiving and store the cluster management device according to it is described request or actively send
Service server network access address, the network access address of the service server is can be for service customer end be provided
The network access address of service;
Address determination module, objective network reference address is defined as one of the described network access address that will be stored;
Connection establishment module, sets up for service server corresponding with the objective network reference address and connects,
Wherein, based on the network access address list from cluster management device, in situation about being forwarded without cluster management device
Under, service customer end directly sets up network connection with the service server to be accessed, and is sent with to one of each service server
Service request, so that the Web content to be accessed is obtained,
The service customer end is further used for the network access address for receiving the described second reception and memory module with virtual
Tree construction be managed and store, and coordinated management is carried out to the data in virtual tree construction using Zookeeper technologies.
5. service customer end according to claim 4, it is characterised in that also include:
Judge module, for when to access the cluster that the cluster management device is managed, judging the service customer end sheet
Whether ground stores at least one network access address of the service server, if being judged as NO, operation second sends mould
Block, conversely, then running connection establishment module.
6. service customer end according to claim 4, it is characterised in that the address determination module, is further used for
When service server corresponding with objective network reference address sets up network connection failure, objective network access is redefined
Location.
7. the service customer end according to any one of claim 4 to 6, it is characterised in that in the service customer end pair
During each time of the service server cluster that the cluster management device is managed is accessed, by the way of poll or using based on visitor
Each network access address that the order that family end attribute determines will be stored is defined as the objective network reference address successively.
8. the service customer end according to any one of claim 4 to 6, it is characterised in that:
It is described second receive and memory module be further used for from the cluster management device receive on network access ground
The corresponding service server self performance of location list and/or the relevant information of loading condition;And
The connection establishment module is based further on service server self performance corresponding with the network access address list
And/or one of the network access address that will be stored of the relevant information of loading condition is defined as the objective network access
Location.
9. a kind of cluster management system, it is characterised in that including the service customer end according to any one of claim 4 to 8,
And cluster management device according to any one of claims 1 to 3 and its each service server in the cluster that is managed.
10. a kind of cluster management method, it is characterised in that including:
Cluster management device receives and stores each described service server sent from each service server in the cluster of management
Respective network access address;
Service customer end sends the request of the network access address for obtaining service server to the cluster management device;
The cluster management device according to it is described acquisition service server network access address request to the business client
End sends at least one network access address of its own storage;
The service customer end receives and stores the network access address of the service server that the cluster management device is sent, will
One of network access address of its own storage is defined as objective network reference address, corresponding with the objective network reference address
Service server set up connection,
Also include:
The cluster management device monitors the running state information of each service server, judges to deposit based on the running state information
Whether the service server corresponding to the network access address of storage can provide service and according to judged result to service customer end
To update stored network access address;
After stored network access address is updated, being judged as to service customer end transmission can for the cluster management device
The network access address of service is provided for service customer end,
Wherein, based on the network access address list from cluster management device, in situation about being forwarded without cluster management device
Under, service customer end directly sets up network connection with the service server to be accessed, and is sent with to one of each service server
Service request, so that the Web content to be accessed is obtained,
The network access address of reception is managed and is stored by the cluster management device with virtual tree construction, in the collection
When group's managing device has managed multiple service server clusters, by the network of the service server in different business server cluster
Reference address is stored in the different subtrees in the virtual tree construction, and using Zookeeper technologies to virtual tree knot
Data in structure carry out coordinated management.
11. methods according to claim 10, it is characterised in that:
The running state information also include on service server self performance corresponding with network access address list and/or
The relevant information of loading condition;And
The service customer end is based on service server self performance corresponding with the network access address list and/or load
One of network access address that the relevant information of situation will be stored is defined as the objective network reference address.
A kind of 12. cluster management methods of cluster management device, it is characterised in that including:
Receive and store each described business sent from each service server in the cluster managed by the cluster management device
The respective network access address of server;
At least one stored network access address is sent to service customer end,
Also include:
Monitor the running state information of each service server;
Judge whether service server corresponding to stored network access address can be to based on the running state information
Service customer end provides and services and update stored network access address according to judged result;
Wherein, it is described to specifically include the step of send stored at least one network access address to service customer end:Xiang Ye
Business client sends the network access address for being judged as that service can be provided for service customer end,
Wherein, based on the network access address list from cluster management device, in situation about being forwarded without cluster management device
Under, service customer end directly sets up network connection with the service server to be accessed, and is sent with to one of each service server
Service request, so that the Web content to be accessed is obtained,
The network access address is managed and is stored with virtual tree construction, and multiple has been managed in the cluster management device
During service server cluster, the network access address of the service server in different business server cluster is stored in the void
In different subtrees in the tree construction of plan, and collaboration pipe is carried out to the data in virtual tree construction using Zookeeper technologies
Reason.
The access method of the service customer end in a kind of 13. clusters, it is characterised in that including:
The request of the network access address for obtaining service server is sent to the cluster management device in the cluster;
Receive and store the network access address of the service server that the cluster management device is sent, the network access address
It is the network access address that service can be provided for service customer end;
One of described network access address that will be stored is defined as objective network reference address;
Connection is set up according to objective network reference address service server corresponding with the objective network reference address, from
And cause that service customer end obtains the Web content to be accessed from corresponding service server,
Based on the network access address list from cluster management device, in the case where being forwarded without cluster management device, industry
Business client directly sets up network connection with the service server to be accessed, please to send business to one of each service server
Ask, so that the Web content to be accessed is obtained,
The service customer end is further used for that the network access address is managed and is stored with virtual tree construction, and
Coordinated management is carried out to the data in virtual tree construction using Zookeeper technologies.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210281817.1A CN103581276B (en) | 2012-08-09 | 2012-08-09 | Cluster management device, system, service customer end and correlation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210281817.1A CN103581276B (en) | 2012-08-09 | 2012-08-09 | Cluster management device, system, service customer end and correlation method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103581276A CN103581276A (en) | 2014-02-12 |
CN103581276B true CN103581276B (en) | 2017-06-09 |
Family
ID=50052191
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210281817.1A Expired - Fee Related CN103581276B (en) | 2012-08-09 | 2012-08-09 | Cluster management device, system, service customer end and correlation method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103581276B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105306514A (en) * | 2014-07-30 | 2016-02-03 | 北京信威通信技术股份有限公司 | Audio recording file distributed storage method and audio recording file distributed storage system based on Zookeeper |
CN105978950A (en) * | 2016-04-27 | 2016-09-28 | 努比亚技术有限公司 | Server resource distribution method, apparatus and system thereof |
CN107948275A (en) * | 2017-11-22 | 2018-04-20 | 四川长虹电器股份有限公司 | The method for realizing the expansible High Availabitity in high in the clouds and high concurrent |
CN108924193A (en) * | 2018-06-15 | 2018-11-30 | 广东小天才科技有限公司 | A kind of service deployment method and device based on server |
CN111083244B (en) * | 2018-10-22 | 2022-09-06 | 浙江宇视科技有限公司 | Cluster address allocation method and device |
CN110046123B (en) * | 2018-12-24 | 2023-06-20 | 创新先进技术有限公司 | Method and device for synchronizing connection states of trusted computing clusters |
CN109981810A (en) * | 2019-03-08 | 2019-07-05 | 福建天泉教育科技有限公司 | A kind of method and system of routing address list update |
CN111726388A (en) * | 2019-03-22 | 2020-09-29 | 苏宁易购集团股份有限公司 | Cross-cluster high-availability implementation method, device, system and equipment |
CN110233886B (en) * | 2019-05-30 | 2021-07-20 | 华南理工大学 | High-availability service management system for massive micro-services and implementation method |
CN112153093B (en) * | 2019-06-28 | 2024-04-16 | 北京京东尚科信息技术有限公司 | Cluster-based task scheduling method, device, equipment and readable storage medium |
CN112988321A (en) * | 2019-12-13 | 2021-06-18 | 北京金山云网络技术有限公司 | Container mirror image data management method and device, electronic equipment and storage medium |
CN112055057B (en) * | 2020-08-14 | 2022-09-30 | 青岛海信网络科技股份有限公司 | Method, system and equipment for dynamic expansion of Web system |
CN112953771A (en) * | 2021-02-22 | 2021-06-11 | 安徽兆尹信息科技股份有限公司 | Financial asset management system based on micro-service and decoupling method |
CN113094074B (en) * | 2021-05-07 | 2024-03-19 | 聚好看科技股份有限公司 | Service cluster updating method and device |
CN113364839A (en) * | 2021-05-26 | 2021-09-07 | 武汉虹旭信息技术有限责任公司 | Service calling method, service calling device and zookeeper cluster |
CN116708459A (en) * | 2023-07-20 | 2023-09-05 | 广州东瑞科技有限公司 | Service access automatic switching method, system, computer equipment and storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101098272A (en) * | 2006-06-27 | 2008-01-02 | 腾讯科技(深圳)有限公司 | Seed enquiring method of P2P system and P2P server |
CN101674329A (en) * | 2009-09-27 | 2010-03-17 | 卓望数码技术(深圳)有限公司 | Internet access method and Internet access system |
CN101714987A (en) * | 2009-11-16 | 2010-05-26 | 中国电信股份有限公司 | P2P playing method and system |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2010049876A2 (en) * | 2008-10-28 | 2010-05-06 | Cotendo Ltd | System and method for sharing transparent proxy between isp and cdn |
-
2012
- 2012-08-09 CN CN201210281817.1A patent/CN103581276B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101098272A (en) * | 2006-06-27 | 2008-01-02 | 腾讯科技(深圳)有限公司 | Seed enquiring method of P2P system and P2P server |
CN101674329A (en) * | 2009-09-27 | 2010-03-17 | 卓望数码技术(深圳)有限公司 | Internet access method and Internet access system |
CN101714987A (en) * | 2009-11-16 | 2010-05-26 | 中国电信股份有限公司 | P2P playing method and system |
Also Published As
Publication number | Publication date |
---|---|
CN103581276A (en) | 2014-02-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103581276B (en) | Cluster management device, system, service customer end and correlation method | |
CN103888277B (en) | A kind of gateway disaster-tolerant backup method, device and system | |
CN104823170B (en) | Distributed caching cluster management | |
CN111615066B (en) | Distributed micro-service registration and calling method based on broadcast | |
CN108234191A (en) | The management method and device of cloud computing platform | |
US20130007253A1 (en) | Method, system and corresponding device for load balancing | |
CN102333029B (en) | Routing method in server cluster system | |
CN110351246A (en) | Server cluster system Socket management method and device | |
CN107947960A (en) | The method for pushing and system of configuration information, the method for reseptance and system of configuration information | |
US11206173B2 (en) | High availability on a distributed networking platform | |
CN103607424B (en) | Server connection method and server system | |
US20030177218A1 (en) | Distributed computer system enhancing a protocol service to a highly available service | |
CN102984501A (en) | Network video-recording cluster system | |
CN109783151B (en) | Method and device for rule change | |
CN103475566A (en) | Real-time message exchange platform and distributed cluster establishment method | |
US20090138751A1 (en) | De-centralized nodal failover handling | |
CN106034137A (en) | Intelligent scheduling method for distributed system, and distributed service system | |
CN109688006B (en) | High-performance weblog message distribution method supporting dynamic detection of target cluster | |
CN108228393A (en) | A kind of implementation method of expansible big data High Availabitity | |
CN103731289A (en) | Method for automatic expansion of network server | |
CN112543150B (en) | Dynamic load balancing method based on server control | |
CN113326100A (en) | Cluster management method, device and equipment and computer storage medium | |
CN106209680A (en) | Information processor and method | |
US8077699B2 (en) | Independent message stores and message transport agents | |
CN115065686B (en) | Configuration method, device and system of distributed load balancing system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170609 Termination date: 20170809 |