CN105939389A - Load balancing method and device - Google Patents
Load balancing method and device Download PDFInfo
- Publication number
- CN105939389A CN105939389A CN201610499271.5A CN201610499271A CN105939389A CN 105939389 A CN105939389 A CN 105939389A CN 201610499271 A CN201610499271 A CN 201610499271A CN 105939389 A CN105939389 A CN 105939389A
- Authority
- CN
- China
- Prior art keywords
- node
- algorithm
- task
- leader node
- waiting task
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
Abstract
Embodiments of the invention provide a load balancing method and device. A leader node obtains load capacity information of processing nodes; and the leader node allocates to-be-processed tasks according to the load capacity information. According to the load balancing method and device, load balancing and high availability of each processing node are realized.
Description
Technical field
The present embodiments relate to field of computer technology, particularly relate to a kind of load-balancing method and device.
Background technology
Canal is the instrument of a real-time data synchronization increased income, and resolves based on data base's Incremental Log, carries
Subscribe to and data consumption for incremental data, currently mainly support mysql (relational data library management system
System).The active and standby duplication of mysql realizes principle, from the point of view of upper strata, is divided into three steps: first, and master (leads)
Change be recorded in binary log (binary log) (these records are called binary log event,
Binary log events, can be checked by show binlog events)；Secondly, slave (standby)
The binary log events of master is copied to its relaying daily record (relay log)；Finally, slave (standby)
Reform relaying daily record in event, will change reflection its own data.
Canal is when work realizes, and the interaction protocol of the slave of simulation mysql, pretending oneself is mysql
Slave, sends dump agreement to mysql master；Mysql master receives dump request, starts
Push binary log to slave (namely Canal)；Canal resolves binary log object.
The server pattern of Canal, it is provided that the mode of High Availabitity, can start multiple node, a joint
After point is because of reason failures such as the machine machine of delaying/process extension, the synchronous task in all of mysql storehouse thereon
Can be automatically transferred on another node.
But, inventor finds during realizing the present invention, the high availability mechanism that Canal provides, and appoints
Business distribution is preemptive type, may the task in all of mysql storehouse be all assigned on same node
Performing, other nodes are all idle, can cause that certain node pressure is excessive, the sky of other node resources
Not busy waste.
Therefore, a kind of method of load balancing urgently proposes.
Summary of the invention
The embodiment of the present invention provides a kind of load-balancing method and device, in order to solve user in prior art
What the task distribution of preemptive type caused respectively processes the defect that node load is unbalance, it is achieved that each process node
Load balancing and high availability.
The embodiment of the present invention provides a kind of load-balancing method, including:
Leader node acquisition processes the load capacity information of node；
Described leader node is according to described load capacity information equilibrium assignment waiting task.
The embodiment of the present invention provides a kind of load balancing apparatus, including:
Data obtaining module, processes the load capacity information of node for leader node acquisition；
Task allocating module, waits to locate according to described load capacity information equilibrium assignment for described leader node
Reason task.
The load-balancing method of embodiment of the present invention offer and device, by according to the process node got
Load capacity information equilibrium assignment waiting task, change prior art is carried out task process time,
What the task distribution of preemptive type caused respectively processes the defect that node load is unbalance, it is achieved that all process nodes
Load balancing and high availability, it is to avoid single node pressure is big, the waste of other node resources, improves system
The overall performance of system is with stable.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to reality
Execute the required accompanying drawing used in example or description of the prior art to be briefly described, it should be apparent that under,
Accompanying drawing during face describes is some embodiments of the present invention, for those of ordinary skill in the art,
On the premise of not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the techniqueflow chart of the embodiment of the present application one；
Fig. 1-1 is the exemplary plot of the embodiment of the present application one；
Fig. 1-2 is another exemplary plot of the embodiment of the present application one；
Fig. 2 is the techniqueflow chart of the embodiment of the present application two；
Fig. 3 is the techniqueflow chart of the embodiment of the present application three；
Fig. 4 is the apparatus structure schematic diagram of the embodiment of the present application four.
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearer, below in conjunction with this
Accompanying drawing in bright embodiment, is clearly and completely described the technical scheme in the embodiment of the present invention,
Obviously, described embodiment is a part of embodiment of the present invention rather than whole embodiments.Based on
Embodiment in the present invention, those of ordinary skill in the art are obtained under not making creative work premise
The every other embodiment obtained, broadly falls into the scope of protection of the invention.
The embodiment of the present application is suitable for inclusion in the group system of multiple process node.
Fig. 1 is the techniqueflow chart of the embodiment of the present application one, and in conjunction with Fig. 1, the embodiment of the present application one is born
The method carrying equilibrium, mainly comprises the following steps that
Step S110, leader node acquisition process the load capacity information of node；
Concrete, in step s 110, described leader node (leader) is from data handling system,
Electing in node according to preset rules processing from starting state, its function is, obtains described
In data handling system, other process the load capacity information of node and carry out the distribution of data processing task.
The described herein node that processes refers both under starting state, can carry out the node of task process, follow-up no longer
Repeat.Concrete, described load capacity information includes the number of the described process node being under starting state
Amount, each described process node be presently processing connection quantity, each described process node the fastest
Response time, the history process performance etc. of each described process node can characterize described process node processing
The information of ability.
Step S120, described leader node are according to described load capacity information equilibrium assignment waiting task.
Concrete, in the step s 120, waiting task described in described leader node equilibrium assignment, according to
The described load capacity information of the described process node having been turned on, uses preset algorithm to described pending
Business carries out equilibrium assignment；Wherein, described preset algorithm includes: concordance hash algorithm, random algorithm,
Round-Robin Algorithm, weighted round-robin algorithm, dynamic Round-Robin Algorithm, the fastest algorithm, minimum join algorithm, anticipation
Combination in any in algorithm.
Wherein, described random algorithm specifically, randomly load is assigned on each available process node,
Choose one by Generating Random Number and process node for task process.
Wherein, described Round-Robin Algorithm is, in order each new connection request is distributed to next process and saves
Point, finally divides all requests to all of process node equally.
In described weighted round-robin algorithm, the connection quantity that each machine accepts is distributed by weight proportion.This
It is the improvement to above-mentioned Round-Robin Algorithm, such as, when a disposal ability processing node is that another processes node
Twice, then described leader node can be distributed to described another the connection quantity of twice and process node.
Described dynamic polling, is similar to above-mentioned weighted round ring, and its difference is, the weight of dynamic repeating query
It is worth based on the lasting monitoring acquisition that each is processed node, and is to constantly update.This is one and moves
State load-balancing algorithm, connects based on the Analyzing Real-Time Performance distribution processing node, the most each node
Current connection number or the fastest response time etc. of node.
The fastest described algorithm distributes based on the fastest response time in all process nodes and connects, and is particularly suited for
Process node in the environment of heterogeneous networks.
Described minimum join algorithm is to process to connect by new task to distribute to the place that current linking number is minimum
Reason node.
In addition to above-mentioned algorithm, in the embodiment of the present application, described preset algorithm can also is that above-mentioned many algorithms
Be used in combination, such as, observe algorithm, utilize Smallest connection algorithm and the fastest algorithm to implement to bear simultaneously
Carry equilibrium.Processing node and obtain a mark according to current connection number and fastest response time, mark is relatively
Gao represents better performances, can more be connected.
The described preset algorithm of the embodiment of the present application can also include that described anticipation algorithm, described anticipation algorithm make
Calculating mark with above-mentioned observation algorithm, the variation tendency analyzing mark judges the property of a certain process node
Can improve or reduce.The process node with improvement trend can more be connected.
In the technical scheme of the embodiment of the present application, when selecting load-balancing algorithm, consistent described in prioritizing selection
Property hash algorithm.
Described concordance hash algorithm (consistent hashing) is the correct algorithm to simple hash algorithm,
It is as follows that it realizes process:
Step S121, by abstract for the overall calculation ability that processes node place system be a ring, to each
Described process node carries out cryptographic Hash for the first time and calculates, and according to the cryptographic Hash obtained by each described process
Node maps on described ring.
Such as 1 in Fig. 1-1,2,3 three nodes it can be understood that be three can perform the place that task processes
Reason node；
Step S122, described waiting task is carried out the described cryptographic Hash of second time calculate, and according to obtaining
Cryptographic Hash, map to described waiting task in the direction of the clock connect most with its cryptographic Hash on described ring
Near node is up.
Such as, in the embodiment of the present application, the cryptographic Hash result of calculation of a certain waiting task and described ring are located
The cryptographic Hash of reason node 3 is closest, then map to this waiting task process on node 3, by
Reason node 3 performs this waiting task.
When processing number of nodes and being less, node is distributed uniform not on ring, and waiting task can be caused to exist
The load capacity of each process node can not be taken into full account during distribution.The embodiment of the present application is to optimize further
Load balancing, is introduced as processing node and arranges the dummy node of correspondence, and each process node can be corresponding
Several dummy nodes.
As shown in the figure of Fig. 1-1, process node 1,2,3,4 skewness.Assuming that four process
Node model is identical, it is assumed that the present load amount processing node 4 is minimum, processes current 3 load capacity of node
Maximum.For load balancing, can be to process node 4 to introduce dummy node 4 " ', as shown in Figure 1-2.
According to the method in step S122, waiting task distribution is processed on node to five, dummy node
4 " ' the waiting task that is assigned to actually distributes in physical processing node 4.So, in certain journey
The load balancing of four physical processing node originally is achieved on degree.
In this enforcement, by pending according to the load capacity information equilibrium assignment processing node got
Task, changes and carries out task in prior art when processing, each process that the task distribution of preemptive type causes
The defect that node load is unbalance, it is achieved that the load balancing of all process nodes and high availability, it is to avoid
Single node pressure is big, the waste of other node resources, improves the overall performance of system and stablizes.
Fig. 2 is the technology six layers figure of the embodiment of the present application two, and in conjunction with Fig. 2, the embodiment of the present application one loads
The method of equilibrium, it is also possible to have a following embodiment:
Step S210, processing in node described at least two, using preset rules to choose leader node；
Concrete, in the embodiment of the present application, choosing of described leader node can be carried out by ZooKeeper
Choose.
ZooKeeper operates in group system, is a distributed application program coordination service, with
Based on Fast Paxos algorithm, it is achieved synchronous service, configuring maintenance and the application of naming service distributed.
As a example by Canal instrument, carry out described leader node when choosing, first in ZooKeeper catalogue
The subdirectory of SEQUENCE, EPHEMERAL type is created under/Canal/nodes
/ Canal/nodes/n_, SEQUENCE type can make ZooKeeper automatically subdirectory name followed by
Upper numbering suffix, more than all numberings existed, (such as, the subdirectory of establishment is this numbering
/Canal/nodes/n_0000000003).EPHEMERAL type can make node in cluster with
After ZooKeeper disconnects, the subdirectory oneself created is automatically deleted.Node reads
ZooKeeper /this catalogue of Canal/nodes/ under all subdirectories, if it find that what oneself was created
Numbering is minimum, and oneself is led exactly.If oneself not being minimum, just in the maximum less than oneself
Subdirectory on arrange watcher monitor, if receive watcher notice, illustrate monitoring that catalogue
Client the most in the dust, oneself re-read all subdirectories, rejudge whether oneself becomes neck
Lead.
It should be noted that in the embodiment of the present application, described leader node is not fixing, if current
Described leader node occurs abnormal, such as delay machine or process exception, and the most remaining process node can be according to
Described preset rules re-elects out a new leader node, thus ensure that the follow-up load of system is equal
Weighing apparatus.
Step S220, described leader node acquisition process the load capacity information of node；
Step S230, described leader node are according to described load capacity information equilibrium assignment waiting task.
Step S220～step S230 are with described in embodiment one, and here is omitted.
The result of described equilibrium assignment is preserved to predetermined directory by step S240, described leader node；
Described predetermined directory is preserved the allocation result of described waiting task, i.e. each described process is saved
Put the corresponding relation with target waiting task.Such as, when Canal carries out data syn-chronization, by right
The acquisition data to be synchronized that resolve of Incremental Log are M, and M is distributed to 4 process by described leader node
Node, it is assumed that M1 is distributed to the first process node and processes, M2 is distributed to the second process
Node carries out processing, M3 distributes to the 3rd process node carries out processing, M4 distributes to the everywhere
Reason node processes, then will preserve M1～M4 and aforementioned four node in described predetermined directory
Corresponding relation.Such as, in ZooKeeper, described predetermined directory can be/Canal/node2instances.
Step S250, described process node monitor described predetermined directory, obtain the target that need to perform pending
Task.
In the embodiment of the present application, described predetermined directory is preserved the target that each described process node is allocated
Process task, described process node (such as, monitors following ZooKeeper by monitoring described predetermined directory
Catalogue :/Canal/node2instances) just can obtain task allocation result, appoint obtaining target process
After business, start worker thread and perform task process.
Execution result according to step S220～step S230, the task allocation result in described predetermined directory
Fill the load capacity of each process node in the system of considering, it is to avoid single process node pressure is big, it
He processes the situation of node resource waste.
Preferably, before step 250, the embodiment of the present application further comprises step S260.
Step S260, for each described target waiting task lock corresponding described process node, thus
Multiple process node is avoided to repeat same described target waiting task.
When system includes multiple process node, for same waiting task, it is possible that multiple
Process node and perform the sight of same waiting task, so can cause repeating to process, waste node resource.
Therefore, use task lock, a target waiting task is locked one and processes node processes process, can
To avoid drawbacks described above well.
Concrete, the described worker thread of each process node, before performance objective task processes, can pass through
Following ZooKeeper catalogue: { instanceName} obtains task lock to/Canal/locks/ $, it is to avoid multinode
Repeat to process.
In the present embodiment, the allocation result of waiting task is obtained by each process node by monitoring leader node
Take the target waiting task of allocated execution, it is not necessary to manual intervention can be realized as multiple process node
Load balancing, it is ensured that the High Availabitity of system and high efficiency；Meanwhile, task lock mechanism is well
Avoid the probability that multiple spot repeats to process, improve the disposal ability of system further.
Fig. 3 is the techniqueflow chart of the embodiment of the present application three, in conjunction with Fig. 3, the embodiment of the present application also include as
Under feasible embodiment:
Step S310, processing in node described at least two, using preset rules to choose leader node；
Step S320, described leader node acquisition process the load capacity information of node；
Step S330, described leader node are according to described load capacity information equilibrium assignment waiting task.
The result of described equilibrium assignment is preserved to predetermined directory by step S340, described leader node；
Step S350, described process node monitor described predetermined directory, obtain the target that need to perform pending
Task；
Step S360, described leader node are when listening to the change of described process node, after change
Described process node equilibrium assignment again described in waiting task to update described predetermined directory.
In this step, described leader node can pass through ZooKeeper in fact to processing monitoring in real time of node
Existing.ZooKeeper supports the concept of watch (observation), can arrange one with system on each process node
Individual observation.If the process node of observed system has change (connect disconnect), then watch will be by
Triggering, the leader node belonging to watch will receive a notification package and be apprised of a certain node and have occurred and that change
Change.
After the process node of system changes (probably due to delay, machine or process hang), part belongs to originally
Process task in the node changed will be unable to continue executing with, and leader node listens to this extremely,
The equilibrium of waiting task can be re-started according to remaining process node according to step S320～step S330
Distribution.
Under alternatively possible situation, if newly-increased dry-cure node in system, in order to the most idle these increase newly
Processing the process resource of node and improve system effectiveness, described leader joint also can be according to step S320～step
S330 re-starts the equilibrium assignment of waiting task to realize preferably load according to remaining process node
Equilibrium.
In the present embodiment, monitored in real time processing node by leader node, and add new in system
Process node, process node and delay machine or process exception exits etc. after reason hangs, redistribute pending
Business, it is not necessary to manual intervention, automatically achieves new load balancing, further increases the availability of system,
Ensure that efficient data process.
Fig. 4 is the apparatus structure schematic diagram of a kind of load balancing apparatus of the embodiment of the present application, in conjunction with Fig. 4,
Described device includes: data obtaining module 41, task allocating module 42, task lock module 43, renewal
Module 44, leader node choose module 45.
Described data obtaining module 41, processes the load capacity information of node for leader node acquisition；
Described task allocating module 42, for described leader node according to the equilibrium point of described load capacity information
Join waiting task.
Further, described task allocating module 42 is additionally operable to: the result of described equilibrium assignment is deposited to
Predetermined directory；Wherein, described predetermined directory is monitored to obtain by described process node and performs described equilibrium
Target waiting task in the result of distribution.
Further, described device also includes that task lock module 43, described task lock module 43 are used for,
Before described acquisition the target waiting task in performing the result of described equilibrium assignment, for each described
Target waiting task locks corresponding described process node, thus avoids multiple process node to repeat
Same described target waiting task.
Further, described device also includes that more new module 44, described more new module 44 are used for: described neck
Lead node when listening to the change of described process node, the most equal according to the described process node after change
Weighing apparatus distributes described waiting task to update described predetermined directory.
Further, described task allocating module 42 is used for: described leader node is according to have been turned on
Process the described load capacity information of node, use preset algorithm that described waiting task carries out equilibrium point
Join；Wherein, described preset algorithm includes: concordance hash algorithm, random algorithm, Round-Robin Algorithm, add
Power Round-Robin Algorithm, dynamic Round-Robin Algorithm, the fastest algorithm, minimum join algorithm, anticipation algorithm in any
Combination.
Further, described device also includes that leader node chooses module 45, and described leader node chooses mould
35 pieces are used for: processing in node described at least two, using preset rules to choose described leader node.
Further, described leader node is chosen module 45 and is additionally operable to: when described leader node exception,
Described preset rules is used to select new described leader node.
In this enforcement, described task allocating module is by the process got according to described data obtaining module
The load capacity information equilibrium assignment waiting task of node, changes and carries out task process in prior art
Time, what the task distribution of preemptive type caused respectively processes the defect that node load is unbalance, it is achieved that all process
The load balancing of node and high availability, it is to avoid single node pressure is big, the waste of other node resources, carries
The overall performance of high system is with stable.
Device embodiment described above is only schematically, wherein said illustrates as separating component
Unit can be or may not be physically separate, the parts shown as unit can be or
Person may not be physical location, i.e. may be located at a place, or can also be distributed to multiple network
On unit.Some or all of module therein can be selected according to the actual needs to realize the present embodiment
The purpose of scheme.Those of ordinary skill in the art are not in the case of paying performing creative labour, the most permissible
Understand and implement.
Through the above description of the embodiments, those skilled in the art is it can be understood that arrive each reality
The mode of executing can add the mode of required general hardware platform by software and realize, naturally it is also possible to by firmly
Part.Based on such understanding, the portion that prior art is contributed by technique scheme the most in other words
Dividing and can embody with the form of software product, this computer software product can be stored in computer can
Read in storage medium, such as ROM/RAM, magnetic disc, CD etc., including some instructions with so that one
Computer installation (can be personal computer, server, or network equipment etc.) performs each to be implemented
The method described in some part of example or embodiment.
Last it is noted that above example is only in order to illustrate technical scheme, rather than to it
Limit；Although the present invention being described in detail with reference to previous embodiment, the ordinary skill of this area
Personnel it is understood that the technical scheme described in foregoing embodiments still can be modified by it, or
Person carries out equivalent to wherein portion of techniques feature；And these amendments or replacement, do not make corresponding skill
The essence of art scheme departs from the spirit and scope of various embodiments of the present invention technical scheme.
Claims (14)
1. the method for a load balancing, it is characterised in that including:
Leader node acquisition processes the load capacity information of node；
Described leader node is according to described load capacity information equilibrium assignment waiting task.
Method the most according to claim 1, it is characterised in that described method also includes:
The result of described equilibrium assignment is deposited to predetermined directory；Wherein, described predetermined directory is by described process
Node is monitored to obtain and performs the target waiting task in the result of described equilibrium assignment.
Method the most according to claim 2, it is characterised in that described acquisition and perform described all
Before target waiting task in the result of weighing apparatus distribution, described method also includes:
Lock corresponding described process node for each described target waiting task, thus avoid multiple process
Node repeats same described target waiting task.
Method the most according to claim 2, it is characterised in that described method also includes:
Described leader node is when listening to the change of described process node, according to the described process joint after change
Described in some equilibrium assignment again, waiting task is to update described predetermined directory.
5. according to the method described in claim 1 or 4, it is characterised in that treat described in described equilibrium assignment
Process task, including:
Described leader node, according to the described load capacity information of the described process node having been turned on, uses and presets
Algorithm carries out equilibrium assignment to described waiting task；Wherein, described preset algorithm includes: concordance is breathed out
Uncommon algorithm, random algorithm, Round-Robin Algorithm, weighted round-robin algorithm, dynamic Round-Robin Algorithm, the fastest algorithm,
Combination in any in minimum join algorithm, anticipation algorithm.
Method the most according to claim 1, it is characterised in that described leader node is at least two
In individual described process node, preset rules is used to choose.
Method the most according to claim 6, it is characterised in that when described leader node exception,
Described preset rules is used to select new described leader node.
8. the device of a load balancing, it is characterised in that including:
Data obtaining module, processes the load capacity information of node for leader node acquisition；
Task allocating module, pending according to described load capacity information equilibrium assignment for described leader node
Task.
Device the most according to claim 8, it is characterised in that described task allocating module is additionally operable to:
The result of described equilibrium assignment is deposited to predetermined directory；Wherein, described predetermined directory is by described process
Node is monitored to obtain and performs the target waiting task in the result of described equilibrium assignment.
Device the most according to claim 9, it is characterised in that described device also includes task locked mode
Block, described task lock module is used for,
Described acquisition and perform described equilibrium assignment result in target waiting task before, for each
Described target waiting task locks corresponding described process node, thus avoids multiple process node to repeat
Perform same described target waiting task.
11. devices according to claim 9, it is characterised in that described device also includes more new module,
Described more new module is used for:
Described leader node is when listening to the change of described process node, according to the described process joint after change
Described in some equilibrium assignment again, waiting task is to update described predetermined directory.
Device described in 12. according to Claim 8 or 11, it is characterised in that described task allocating module
For:
Described leader node, according to the described load capacity information of the described process node having been turned on, uses and presets
Algorithm carries out equilibrium assignment to described waiting task；Wherein, described preset algorithm includes: concordance is breathed out
Uncommon algorithm, random algorithm, Round-Robin Algorithm, weighted round-robin algorithm, dynamic Round-Robin Algorithm, the fastest algorithm,
Combination in any in minimum join algorithm, anticipation algorithm.
13. devices according to claim 8, it is characterised in that described device also includes, Ling Daojie
Click delivery block, described leader node choose module for:
Processing in node described at least two, using preset rules to choose described leader node.
14. devices according to claim 13, it is characterised in that described leader node chooses module
It is additionally operable to:
When described leader node exception, described preset rules is used to select new described leader node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610499271.5A CN105939389A (en) | 2016-06-29 | 2016-06-29 | Load balancing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610499271.5A CN105939389A (en) | 2016-06-29 | 2016-06-29 | Load balancing method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105939389A true CN105939389A (en) | 2016-09-14 |
Family
ID=56873059
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610499271.5A Pending CN105939389A (en) | 2016-06-29 | 2016-06-29 | Load balancing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105939389A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106921754A (en) * | 2017-05-04 | 2017-07-04 | 泰康保险集团股份有限公司 | The load-balancing method of group system, device, medium and electronic equipment |
| CN107360208A (en) * | 2017-05-27 | 2017-11-17 | 武汉斗鱼网络科技有限公司 | Promote method, system, storage medium and the service cluster of task status |
| CN108268314A (en) * | 2016-12-31 | 2018-07-10 | 北京亿阳信通科技有限公司 | A kind of method of multithreading task concurrent processing |
| CN109062681A (en) * | 2018-08-24 | 2018-12-21 | 郑州云海信息技术有限公司 | A kind of execution method, system, device and the storage medium of periodic cycle task |
| CN109144783A (en) * | 2018-08-22 | 2019-01-04 | 南京壹进制信息技术股份有限公司 | A kind of distribution magnanimity unstructured data backup method and system |
| CN109150758A (en) * | 2017-06-19 | 2019-01-04 | 中兴通讯股份有限公司 | Node traffic distribution method, device, system and computer readable storage medium |
| WO2019061122A1 (en) * | 2017-09-28 | 2019-04-04 | 深圳大学 | Spark task allocation method and system |
| CN109587220A (en) * | 2018-11-09 | 2019-04-05 | 金蝶软件（中国）有限公司 | Load-balancing method, device, computer equipment and storage medium |
| CN109815014A (en) * | 2019-01-17 | 2019-05-28 | 北京三快在线科技有限公司 | Data processing method, device, electronic equipment and computer readable storage medium |
| CN109829003A (en) * | 2018-12-25 | 2019-05-31 | 北京市天元网络技术股份有限公司 | Database backup method and device |
| CN110049091A (en) * | 2019-01-10 | 2019-07-23 | 阿里巴巴集团控股有限公司 | Date storage method and device, electronic equipment, storage medium |
| CN111124657A (en) * | 2018-10-31 | 2020-05-08 | 北京金山云网络技术有限公司 | Resource management method, device, electronic equipment and storage medium |
| WO2020147330A1 (en) * | 2019-01-18 | 2020-07-23 | 苏宁云计算有限公司 | Data stream processing method and system |
| CN116155910A (en) * | 2023-03-29 | 2023-05-23 | 新华三技术有限公司 | Equipment management method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102081554A (en) * | 2011-01-30 | 2011-06-01 | 浪潮(北京)电子信息产业有限公司 | Cloud computing operating system as well as kernel control system and method thereof |
| CN102185779A (en) * | 2011-05-11 | 2011-09-14 | 田文洪 | Method and device for realizing data center resource load balance in proportion to comprehensive allocation capability |
| CN102577241A (en) * | 2009-12-31 | 2012-07-11 | 华为技术有限公司 | Method, device and system for scheduling distributed buffer resources |
| CN103259832A (en) * | 2012-12-24 | 2013-08-21 | 中国科学院沈阳自动化研究所 | Cluster resource control method for achieving dynamic load balance, fault diagnosis and failover |
| US20160057214A1 (en) * | 2013-04-15 | 2016-02-25 | P. Ashok Anand | Load Balancer System and Method for Server/Nodes in Cloud Environment |
-
2016
- 2016-06-29 CN CN201610499271.5A patent/CN105939389A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102577241A (en) * | 2009-12-31 | 2012-07-11 | 华为技术有限公司 | Method, device and system for scheduling distributed buffer resources |
| CN102081554A (en) * | 2011-01-30 | 2011-06-01 | 浪潮(北京)电子信息产业有限公司 | Cloud computing operating system as well as kernel control system and method thereof |
| CN102185779A (en) * | 2011-05-11 | 2011-09-14 | 田文洪 | Method and device for realizing data center resource load balance in proportion to comprehensive allocation capability |
| CN103259832A (en) * | 2012-12-24 | 2013-08-21 | 中国科学院沈阳自动化研究所 | Cluster resource control method for achieving dynamic load balance, fault diagnosis and failover |
| US20160057214A1 (en) * | 2013-04-15 | 2016-02-25 | P. Ashok Anand | Load Balancer System and Method for Server/Nodes in Cloud Environment |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108268314A (en) * | 2016-12-31 | 2018-07-10 | 北京亿阳信通科技有限公司 | A kind of method of multithreading task concurrent processing |
| CN106921754B (en) * | 2017-05-04 | 2020-07-28 | 泰康保险集团股份有限公司 | Load balancing method, device, medium and electronic equipment of cluster system |
| CN106921754A (en) * | 2017-05-04 | 2017-07-04 | 泰康保险集团股份有限公司 | The load-balancing method of group system, device, medium and electronic equipment |
| CN107360208B (en) * | 2017-05-27 | 2019-11-01 | 武汉斗鱼网络科技有限公司 | Promote method, system, storage medium and the service cluster of task status |
| CN107360208A (en) * | 2017-05-27 | 2017-11-17 | 武汉斗鱼网络科技有限公司 | Promote method, system, storage medium and the service cluster of task status |
| CN109150758A (en) * | 2017-06-19 | 2019-01-04 | 中兴通讯股份有限公司 | Node traffic distribution method, device, system and computer readable storage medium |
| WO2019061122A1 (en) * | 2017-09-28 | 2019-04-04 | 深圳大学 | Spark task allocation method and system |
| CN109144783A (en) * | 2018-08-22 | 2019-01-04 | 南京壹进制信息技术股份有限公司 | A kind of distribution magnanimity unstructured data backup method and system |
| CN109144783B (en) * | 2018-08-22 | 2020-08-18 | 南京壹进制信息科技有限公司 | Distributed massive unstructured data backup method and system |
| CN109062681A (en) * | 2018-08-24 | 2018-12-21 | 郑州云海信息技术有限公司 | A kind of execution method, system, device and the storage medium of periodic cycle task |
| CN111124657A (en) * | 2018-10-31 | 2020-05-08 | 北京金山云网络技术有限公司 | Resource management method, device, electronic equipment and storage medium |
| CN111124657B (en) * | 2018-10-31 | 2024-03-08 | 北京金山云网络技术有限公司 | Resource management method, device, electronic equipment and storage medium |
| CN109587220A (en) * | 2018-11-09 | 2019-04-05 | 金蝶软件（中国）有限公司 | Load-balancing method, device, computer equipment and storage medium |
| CN109587220B (en) * | 2018-11-09 | 2021-09-10 | 金蝶软件（中国）有限公司 | Load balancing method and device, computer equipment and storage medium |
| CN109829003A (en) * | 2018-12-25 | 2019-05-31 | 北京市天元网络技术股份有限公司 | Database backup method and device |
| CN110049091A (en) * | 2019-01-10 | 2019-07-23 | 阿里巴巴集团控股有限公司 | Date storage method and device, electronic equipment, storage medium |
| CN109815014A (en) * | 2019-01-17 | 2019-05-28 | 北京三快在线科技有限公司 | Data processing method, device, electronic equipment and computer readable storage medium |
| WO2020147330A1 (en) * | 2019-01-18 | 2020-07-23 | 苏宁云计算有限公司 | Data stream processing method and system |
| CN116155910A (en) * | 2023-03-29 | 2023-05-23 | 新华三技术有限公司 | Equipment management method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105939389A (en) | Load balancing method and device | |
| EP3522013B1 (en) | Method and system for migration of containers in a container orchestration platform between compute nodes | |
| CN106844510B (en) | Data migration method and device for distributed database cluster | |
| CN107959705B (en) | Distribution method of streaming computing task and control server | |
| CN110737442A (en) | edge application management method and system | |
| US20190075084A1 (en) | Distributed Lock Management Method, Apparatus, and System | |
| CN111880936B (en) | Resource scheduling method, device, container cluster, computer equipment and storage medium | |
| CN108881512B (en) | CTDB virtual IP balance distribution method, device, equipment and medium | |
| CN101984632A (en) | Load distributing method, device and server in distributed cache system | |
| CN105471960A (en) | Information interaction system and method between private clouds and public cloud | |
| CN110244901B (en) | Task allocation method and device and distributed storage system | |
| CN111427670A (en) | Task scheduling method and system | |
| CN112463395A (en) | Resource allocation method, device, equipment and readable storage medium | |
| CN105827678A (en) | High-availability framework based communication method and node | |
| CN111092754B (en) | Real-time access service system and implementation method thereof | |
| CN112468310B (en) | Streaming media cluster node management method and device and storage medium | |
| CN114338670B (en) | Edge cloud platform and network-connected traffic three-level cloud control platform with same | |
| CN111866210A (en) | Virtual IP balance distribution method, system, terminal and storage medium | |
| CN112395313A (en) | Global sequence generation method and device, electronic equipment and computer storage medium | |
| EP3648423B1 (en) | Resource allocation method and system | |
| CN109327520B (en) | Method and device for establishing connection between terminal and server node | |
| CN115145782A (en) | Server switching method, mooseFS system and storage medium | |
| CN115687250A (en) | Storage method, equipment, system and computer storage medium | |
| WO2022220830A1 (en) | Geographically dispersed hybrid cloud cluster | |
| JP2013182553A (en) | Management device and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160914 |