CN102571965A - Method and device for starting service group in high-availability cluster - Google Patents
Method and device for starting service group in high-availability cluster Download PDFInfo
- Publication number
- CN102571965A CN102571965A CN2012100128330A CN201210012833A CN102571965A CN 102571965 A CN102571965 A CN 102571965A CN 2012100128330 A CN2012100128330 A CN 2012100128330A CN 201210012833 A CN201210012833 A CN 201210012833A CN 102571965 A CN102571965 A CN 102571965A
- Authority
- CN
- China
- Prior art keywords
- service groups
- server
- conflict
- service
- starts
- 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.)
- Granted
Links
Images
Landscapes
- Stored Programmes (AREA)
Abstract
The invention discloses a method and a device for starting a service group in a high-availability cluster. The method comprises the following step of: when a certain service group is to be started, if the service group can be started on a plurality of servers, and has mutually exclusive conflict service groups, starting the service group after a maximum number of mutually exclusive conflict service groups are started. Finite server resources in the cluster are combined, a maximum number of service groups are started by utilizing a maximal starting algorithm according to mutually exclusive conflict relationship between every two service groups in the cluster, and then the service groups are optimally started according to different priorities of the service group on the servers, so that a maximum number of services can be provided by the cluster, and the maximum performance of the cluster is also ensured.
Description
Technical field
The present invention relates to the high available Clustering of computer, relate in particular to method and the device of realizing that service groups starts in the high available cluster.
Background technology
At present, along with the arrival in cloud epoch, the application of application, the especially highly available cluster system of more computer clusters of stressing of forming by large-scale computer more.
Common form with service groups provides the multiple business service in a highly available cluster system, and these service groups possibly exist the mutual exclusion conflict relationship.Exist a plurality of service groups of mutual exclusion conflict relationship to be called mutual exclusion conflict service groups, these service groups can not be enabled on the server simultaneously, to prevent them since between repel each other and service can not normally be provided.In cluster was provided with, the first Priority Service device of two mutual exclusion conflict service groups was inequality.Cause thus in the limited server resource of cluster, cause the service groups in the cluster service can not normally start the service that provides best because of Starting mode is improper.
In the available cluster service of existing height, the startup of mutual exclusion conflict service groups adopts competitive mode to start basically, promptly starts to afoul service groups on the server of optimum (resource and operation etc.).And if certain service groups does not have the startup server, then have to start.Will cause some service groups that service can not normally externally be provided like this.
Therefore, need provide and realize the method that service groups starts in a kind of high available cluster, can start the service groups in the cluster as much as possible, under feasible condition, can service groups be started on the highest server of bootable priority simultaneously.
Summary of the invention
Technical problem to be solved by this invention provides a kind of method and device of realizing that service groups starts in the high available cluster, can start the service groups in the cluster as much as possible.
In order to solve the problems of the technologies described above, the invention provides a kind of method that realizes that service groups starts in the high available cluster, comprising:
When having service groups to start, have a plurality of bootable servers if judge this service groups, and there is mutual exclusion conflict service groups in this service groups, then after starting mutual exclusion conflict service groups as much as possible, start this service groups.
Further,
If judge that there are a plurality of bootable servers in this service groups; And there is mutual exclusion conflict service groups in this service groups; Then, start this service groups through the optimization starting algorithm through maximizing starting algorithm and/or optimization starting algorithm after starting mutual exclusion conflict service groups as much as possible;
Wherein, the maximization starting algorithm is meant from the bootable server of one or more mutual exclusions conflict service groups that this service groups exists, and picks out the service groups of conflicting with the corresponding mutual exclusion of the nonoverlapping startup of server of the bootable server of this service groups; The optimization starting algorithm starts mutual exclusion conflict service groups or the corresponding service group is meant: from a plurality of start to find out the server start the highest server of said mutual exclusion conflict service groups or corresponding service group priority or start at unique server that starts, then from starting the record that deletion the server record has started server.
Further, this method also comprises:
Have a plurality of bootable servers if judge this service groups, and there is not mutual exclusion conflict service groups in this service groups, then directly starts this service groups through the optimization starting algorithm.
Further, this method also comprises:
If judge that only there is a bootable server in this service groups, then directly on this server, start this service groups.
In order to solve the problems of the technologies described above, the invention provides a kind of device of realizing that service groups starts in the high available cluster, comprise that the service groups that connects successively starts module, the equipment that can open searches module and the conflict service groups is searched module, wherein:
The equipment that can open is searched module, is used for when having service groups to start, if finding this service groups exists a plurality of bootable servers, then searches the look-up command of module output conflict service groups to the conflict service groups; If the conflict service groups is searched module and returned corresponding mutual exclusion conflict service groups, then start the enabled instruction of module output conflict service groups to service groups;
The conflict service groups is searched module, is used for according to the conflict service groups look-up command of input the mutual exclusion conflict service groups of the corresponding service group that finds being searched module to the equipment that can open and returns;
Service groups starts module, is used for based on the conflict service groups enabled instruction of importing after starting mutual exclusion conflict service groups as much as possible, starting this service groups.
Further,
Service groups starts module and starts mutual exclusion conflict service groups as much as possible through the optimization starting algorithm, and starts this service groups through the optimization starting algorithm; Wherein, the maximization starting algorithm is meant: from the bootable server of one or more mutual exclusions conflict service groups that this service groups exists, pick out with this service groups and can start the corresponding mutual exclusion of the nonoverlapping startup of server of the server service groups of conflicting; The optimization starting algorithm starts mutual exclusion conflict service groups or the corresponding service group is meant: from a plurality of start to find out the server start the highest server of mutual exclusion conflict service groups or corresponding service group priority or start at unique server that starts, then from starting the record that deletion the server record has started server.
Further,
The conflict service groups is searched module according to the look-up command of said conflict service groups, does not exist the lookup result of mutual exclusion conflict service groups to return to the equipment that can open the corresponding service group and searches module;
The equipment that can open is searched module and is not had the lookup result of mutual exclusion conflict service groups according to the corresponding service group, starts module output optimization to service groups and starts the service groups instruction;
Service groups starts module and starts the corresponding business group based on the optimization startup service groups instruction of importing through the optimization starting algorithm.
Further,
The equipment that can open is searched module if searching this service groups only exists a bootable server, then starts module output to service groups and on given server, starts the service groups instruction;
Said service groups starts module according on given server, starting the service groups instruction, directly on given server, starts the corresponding business group.
The present invention combines server resource limited in the cluster, and according to the mutual exclusion conflict relationship between the service groups in the cluster, at first utilizes the maximization starting algorithm to start service groups as much as possible; On this basis, the different service groups optimizations that realize of the priority of utilizing service groups on the period of service, to start start.So both guaranteed that cluster provided maximum business service, also guaranteed cluster performance maximum efficiency simultaneously.
Description of drawings
Fig. 1 is the flow chart of the method embodiment that service groups starts in the high available cluster of realization of the present invention;
Fig. 2 is the flow chart of optimization starting algorithm instance in Fig. 1 method embodiment flow process;
Fig. 3 is the structured flowchart of the device embodiment that service groups starts in the high available cluster of realization of the present invention.
Embodiment
Below in conjunction with accompanying drawing and preferred embodiment technical scheme of the present invention is described in detail.Should be appreciated that following listed examples only is used for explanation and explains the present invention, and does not constitute the restriction to technical scheme of the present invention.
The present invention at first judges and should organize existing how many available startup servers when starting certain group; If only have a server, then directly service groups be enabled on this server; If a plurality of servers that start are arranged, then need consider the startup of the mutual exclusion conflict service groups (being called for short the conflict service groups) of this service groups, the service groups if this service groups is not conflicted then can directly utilize the optimization starting algorithm to start this service groups; If there is the conflict service groups, then after taking into full account and mutual exclusion conflict service groups as much as possible started, start this service groups, realize that maximization starts.
The method embodiment that service groups starts in the high available cluster of realization of the present invention comprises:
When having service groups to start, have a plurality of bootable servers if judge this service groups, and there is mutual exclusion conflict service groups in this service groups, then after starting mutual exclusion conflict service groups as much as possible, start this service groups.
In said method embodiment,
Start mutual exclusion conflict service groups as much as possible through maximization starting algorithm and/or optimization starting algorithm, and start this service groups through the optimization starting algorithm;
Wherein, the maximization starting algorithm is meant from the bootable server of one or more mutual exclusions conflict service groups that this service groups exists, and picks out with this service groups to start the corresponding mutual exclusion of the nonoverlapping startup of server of the server service groups of conflicting; The optimization starting algorithm starts mutual exclusion conflict service groups or the corresponding service group is meant: from a plurality of start to find out the server start the highest server of mutual exclusion conflict service groups or corresponding service group priority or start at unique server that starts, then from starting the record that deletion the server record has started server.
In said method embodiment,
Have a plurality of bootable servers if judge this service groups, and there is not mutual exclusion conflict service groups in this service groups, then directly starts this service groups through the optimization starting algorithm.
In said method embodiment,
If judge that only there is a bootable server in this service groups, directly on this server, start this service groups.
Fig. 1 has represented the flow process of the method embodiment that service groups starts in the high available cluster of above-mentioned realization, comprises the steps:
101: have service groups to start;
102: judging to this service groups whether have a plurality of servers that start, is then to carry out next step, otherwise execution in step 114;
103: preserve the record A that this service groups can start server;
104: judging to this service groups whether have mutual exclusion conflict service groups (being called for short the conflict service groups), is then to carry out next step, otherwise execution in step 113;
105: the record of Polling conflict service groups, and obtain one of them conflict service groups;
106: judge whether this conflict service groups exists a plurality of servers that start, and is then to carry out next step, otherwise execution in step 112;
107: preserve the record B that this conflict service groups can start server;
108: judge whether this record B exists other server beyond the record A, be then to return step 105 to carry out, otherwise carry out next step;
109: carry out the optimization starting algorithm and start the service groups of to conflict;
110: the server record that deletion has started from record A;
111: judge that whether bootable number of servers is then return step 105 carry out greater than 1 among the record A, otherwise execution in step 113;
112: on corresponding server, start the service groups of to conflict, execution in step 110;
113: carry out the optimization starting algorithm and start this service groups, process ends;
114: directly on corresponding server, start this service groups, process ends.
Fig. 2 has represented the flow process of step 109 in Fig. 1 method embodiment flow process, 113 execution optimization starting algorithm instances, comprises the steps:
210: judging among the record A whether a plurality of servers that start are arranged, is then to carry out next step, otherwise execution in step 230;
220: find out the highest server of startup corresponding business group priority the server from a plurality of startup of record A;
At this, start the highest server of corresponding business group priority and be meant that server that is suitable for starting operation corresponding service group most of having got rid of interdynamic factor.
230: on the server of selecting, start the corresponding business group;
240: deletion has started the server record of service groups, process ends from record A.
The present invention is directed to said method embodiment, the device embodiment that realizes that service groups starts in the high available cluster correspondingly also is provided, comprise that the service groups that connects successively starts module, the equipment that can open searches module and the conflict service groups is searched module, wherein:
The equipment that can open is searched module, is used for when having service groups to start, if finding this service groups exists a plurality of bootable servers, then searches the look-up command of module output conflict service groups to the conflict service groups; If the conflict service groups is searched module and returned corresponding mutual exclusion conflict service groups, then start the enabled instruction of module output conflict service groups to service groups;
The conflict service groups is searched module, is used for according to the conflict service groups look-up command of input the mutual exclusion conflict service groups of the corresponding service group that finds being searched module to the equipment that can open and returns;
Service groups starts module, is used for conflicting the service groups enabled instruction after starting mutual exclusion conflict service groups as much as possible based on the mutual exclusion of input, starts this service groups.
In said apparatus embodiment,
Service groups starts module and starts mutual exclusion conflict service groups as much as possible through the optimization starting algorithm, and starts this service groups through the optimization starting algorithm; Wherein, the maximization starting algorithm is meant from the bootable server of one or more mutual exclusions conflict service groups that this service groups exists, and picks out with this service groups to start the corresponding mutual exclusion of the nonoverlapping startup of server of the server service groups of conflicting; The optimization starting algorithm starts mutual exclusion conflict service groups or the corresponding service group is meant: from a plurality of start to find out the server start the highest server of mutual exclusion conflict service groups or corresponding service group priority or start at unique server that starts, then from starting the record that deletion the server record has started server.
In said apparatus embodiment,
The conflict service groups is searched the conflict service groups look-up command of module according to input, does not exist the lookup result of mutual exclusion conflict service groups to return to the equipment that can open the corresponding service group and searches module;
The equipment that can open is searched module and is searched the lookup result that there is not mutual exclusion conflict service groups in corresponding service group that module returns according to the conflict service groups, starts module output optimization to service groups and starts the service groups instruction;
Service groups starts module and starts the corresponding business group based on the optimization startup service groups instruction of importing through the optimization starting algorithm.
In said apparatus embodiment,
The equipment that can open is searched module if searching this service groups only exists a bootable server, then starts module output to service groups and on given server, starts the service groups instruction;
Service groups starts module instructs according to the service groups that on given server, starts of input, directly on given server, starts the corresponding business group.
The present invention combines limited server resource, and according to the mutual exclusion conflict relationship between the service groups, at first utilizes the maximization starting algorithm to provide maximization to start the method for service groups; On this basis, the priority of utilizing service groups on the period of service, to start is different, realizes that optimization starts service groups.Guaranteed that so high available set group energy provides business service enough as much as possible, guarantees the maximized usefulness of cluster with this.
To those skilled in the art; After having understood content of the present invention and principle; Can be under the situation that does not deviate from the principle and scope of the present invention; Carry out various corrections and change on form and the details according to the method for the invention, but these are based on correction of the present invention with change still within claim protection range of the present invention.
Claims (8)
1. a method that realizes that service groups starts in the high available cluster is characterized in that, comprising:
When having service groups to start, have a plurality of bootable servers if judge this service groups, and there is mutual exclusion conflict service groups in this service groups, then after starting mutual exclusion conflict service groups as much as possible, start this service groups.
2. according to the described method of claim 1, it is characterized in that,
Said if judge that there are a plurality of bootable servers in this service groups; And there is mutual exclusion conflict service groups in this service groups; Then, start this service groups through said optimization starting algorithm through maximizing starting algorithm and/or optimization starting algorithm after starting mutual exclusion conflict service groups as much as possible;
Wherein, the maximization starting algorithm is meant from the bootable server of one or more mutual exclusions conflict service groups that this service groups exists, and picks out the service groups of conflicting with the corresponding mutual exclusion of the nonoverlapping startup of server of the bootable server of this service groups; The optimization starting algorithm starts mutual exclusion conflict service groups or the corresponding service group is meant: from a plurality of start to find out the server start the highest server of said mutual exclusion conflict service groups or said corresponding service group priority or carry out said startup at unique server that starts, then from starting the record that deletion the server record has started server.
3. according to the described method of claim 2, it is characterized in that, also comprise:
Have a plurality of bootable servers if judge this service groups, and there is not mutual exclusion conflict service groups in this service groups, then directly through said this service groups of optimization starting algorithm startup.
4. according to the described method of claim 2, it is characterized in that, also comprise:
If judge that only there is a bootable server in this service groups, then directly on this server, start this service groups.
5. a device of realizing that service groups starts in the high available cluster comprises that the service groups that connects successively starts module, the equipment that can open searches module and the conflict service groups is searched module, wherein:
The equipment that can open is searched module, is used for when having service groups to start, if finding this service groups exists a plurality of bootable servers, then searches the look-up command of module output conflict service groups to the conflict service groups; If the conflict service groups is searched module and returned corresponding mutual exclusion conflict service groups, then start the enabled instruction of module output conflict service groups to service groups;
The conflict service groups is searched module, is used for according to the said conflict service groups look-up command of input the mutual exclusion conflict service groups of the corresponding service group that finds being searched module to the equipment that can open and returns;
Service groups starts module, is used for based on the said conflict service groups enabled instruction of importing after starting mutual exclusion conflict service groups as much as possible, starting this service groups.
6. according to the described device of claim 5, it is characterized in that,
Said service groups starts module and starts mutual exclusion conflict service groups as much as possible through the optimization starting algorithm, and starts this service groups through the optimization starting algorithm; Wherein, said maximization starting algorithm is meant from the bootable server of one or more mutual exclusions conflict service groups that this service groups exists, and picks out with this service groups to start the corresponding mutual exclusion of the nonoverlapping startup of server of the server service groups of conflicting; Said optimization starting algorithm starts mutual exclusion conflict service groups or the corresponding service group is meant: from a plurality of start to find out the server start the highest server of mutual exclusion conflict service groups or corresponding service group priority or start at unique server that starts, then from starting the record that deletion the server record has started server.
7. according to the described device of claim 6, it is characterized in that,
Said conflict service groups is searched module according to the look-up command of said conflict service groups, does not exist the lookup result of mutual exclusion conflict service groups to return to the said equipment that opens the corresponding service group and searches module;
The said equipment that opens is searched module and is not had the lookup result of mutual exclusion conflict service groups according to said corresponding service group, starts module output optimization to said service groups and starts the service groups instruction;
Said service groups starts module and instructs through said optimization starting algorithm startup corresponding business group according to the said optimization startup service groups of input.
8. according to the described device of claim 6, it is characterized in that,
The said equipment that opens is searched module if searching this service groups only exists a bootable server, then starts module output to said service groups and on given server, starts the service groups instruction;
Said service groups starts module based on said service groups instruction, directly the startup corresponding business group on given server of on given server, starting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210012833.0A CN102571965B (en) | 2012-01-16 | 2012-01-16 | Method and device for starting service group in high-availability cluster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210012833.0A CN102571965B (en) | 2012-01-16 | 2012-01-16 | Method and device for starting service group in high-availability cluster |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102571965A true CN102571965A (en) | 2012-07-11 |
| CN102571965B CN102571965B (en) | 2014-08-13 |
Family
ID=46416425
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210012833.0A Active CN102571965B (en) | 2012-01-16 | 2012-01-16 | Method and device for starting service group in high-availability cluster |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102571965B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017049997A1 (en) * | 2015-09-25 | 2017-03-30 | 华为技术有限公司 | Virtual machine monitoring method, apparatus and system based on cloud computing service |
| CN109587223A (en) * | 2018-11-20 | 2019-04-05 | 北京奇艺世纪科技有限公司 | Data aggregation method, device and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060129989A1 (en) * | 2004-12-09 | 2006-06-15 | Christian Fleischer | System and method for registering native libraries with non-native enterprise program code |
| US7181574B1 (en) * | 2003-01-30 | 2007-02-20 | Veritas Operating Corporation | Server cluster using informed prefetching |
| CN101699882A (en) * | 2009-10-21 | 2010-04-28 | 深圳华为通信技术有限公司 | Method, device and system for implementing interaction between color ring back tone service and supplementary service |
| CN102238487A (en) * | 2010-05-07 | 2011-11-09 | 北京信威通信技术股份有限公司 | Multimedia cluster scheduling system and method thereof |
| CN102255960A (en) * | 2011-06-28 | 2011-11-23 | 用友软件股份有限公司 | Single-point service processing device and single-point service processing method |
-
2012
- 2012-01-16 CN CN201210012833.0A patent/CN102571965B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7181574B1 (en) * | 2003-01-30 | 2007-02-20 | Veritas Operating Corporation | Server cluster using informed prefetching |
| US20060129989A1 (en) * | 2004-12-09 | 2006-06-15 | Christian Fleischer | System and method for registering native libraries with non-native enterprise program code |
| CN101699882A (en) * | 2009-10-21 | 2010-04-28 | 深圳华为通信技术有限公司 | Method, device and system for implementing interaction between color ring back tone service and supplementary service |
| CN102238487A (en) * | 2010-05-07 | 2011-11-09 | 北京信威通信技术股份有限公司 | Multimedia cluster scheduling system and method thereof |
| CN102255960A (en) * | 2011-06-28 | 2011-11-23 | 用友软件股份有限公司 | Single-point service processing device and single-point service processing method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017049997A1 (en) * | 2015-09-25 | 2017-03-30 | 华为技术有限公司 | Virtual machine monitoring method, apparatus and system based on cloud computing service |
| CN109587223A (en) * | 2018-11-20 | 2019-04-05 | 北京奇艺世纪科技有限公司 | Data aggregation method, device and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102571965B (en) | 2014-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2018017260A1 (en) | Accelerator resource allocation and pooling | |
| CN103379159B (en) | A kind of method that distributed Web station data synchronizes | |
| CN101594319B (en) | Entry lookup method and entry lookup device | |
| CN105933376A (en) | Data manipulation method, server and storage system | |
| CN104158911A (en) | Server cluster-oriented control method and system | |
| CN104820717A (en) | Massive small file storage and management method and system | |
| CN104572274A (en) | Cross-cloud-node migration system and cross-cloud-node migration method | |
| CN102375837A (en) | Data acquiring system and method | |
| WO2010117691A3 (en) | Priority-based management of system load level | |
| CN104184812A (en) | Multi-point data transmission method based on private cloud | |
| CN103164253A (en) | Virtual machine deployment system and virtual machine deployment method | |
| Zhang et al. | Smarth: Enabling multi-pipeline data transfer in hdfs | |
| CN106953910A (en) | A kind of Hadoop calculates storage separation method | |
| CN104750690A (en) | Query processing method, device and system | |
| CN103246549B (en) | A kind of method and system of data conversion storage | |
| WO2017092384A1 (en) | Clustered database distributed storage method and device | |
| CN102142032A (en) | Method and system for reading and writing data of distributed file system | |
| CN104410666A (en) | Method and system for implementing heterogeneous storage resource management under cloud computing | |
| CN103116655A (en) | Clustered data query method, client side and system | |
| CN102904977A (en) | Network address allocation method, server and node | |
| CN102664922A (en) | High-speed network starting method based on Linux system | |
| CN102082832A (en) | Data storage method and device in distributed file system | |
| CN102571965A (en) | Method and device for starting service group in high-availability cluster | |
| CN103116475A (en) | Method of automatic simplifying allocation expansion | |
| CN104778125A (en) | Method and system for memory management |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |