CN109815008A - Hadoop cluster user resource monitoring method and system - Google Patents
Hadoop cluster user resource monitoring method and system Download PDFInfo
- Publication number
- CN109815008A CN109815008A CN201811573182.6A CN201811573182A CN109815008A CN 109815008 A CN109815008 A CN 109815008A CN 201811573182 A CN201811573182 A CN 201811573182A CN 109815008 A CN109815008 A CN 109815008A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- hadoop cluster
- statistical
- resource monitoring
- 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
Landscapes
- Debugging And Monitoring (AREA)
Abstract
This disclosure relates to a kind of Hadoop cluster user resource monitoring method and system, wherein the described method includes: from the job log information of the active and standby node of Hadoop cluster acquisition user；Data parameters from the job log information of user according to setting extract corresponding useful data；And the useful data is counted according to statistical parameter, obtain corresponding statistical data.The disclosure can monitor the resource service condition of each user in cluster in real time, cluster administrator is set accurately to grasp each user to the service condition of resource, and according to user to the service condition of resource, more reasonably available resources is distributed for user, so as to more reasonably adjust cluster resource, dilatation whether is needed for cluster and how dilatation provides foundation.
Description
Technical field
This disclosure relates to Hadoop cluster resource administrative skill field, and in particular, to a kind of Hadoop cluster user money
Source monitoring method and system.
Background technique
Demand currently to Hadoop cluster is more and more extensive, and the rental of Hadoop cluster multi-user is more universal, by dividing
The Yarn of layer architecture is particularly important come the reasonable scheduling of resource under multi-user's task for completing.In the system framework of Yarn
There are two important components: one is Resourcemanager (abbreviation RM), is responsible for the resource scheduling management of task；Another
It is the Applicationmaster (abbreviation AM) for forming one-to-one mode with the application task of user, is responsible for application resource and supervises
Control task run situation.
In RM, the scheduling method of scheduler Scheduler can be used official offer FIFO first in first out mode,
Capacity lining up mode or Fairshare flexibly divide resources mode equally, can also be with customized scheduling rule.Wherein,
It is the mode for using queue that Capacity lining up mode and Fairshare, which flexibly divide resource equally, internal still according to FIFO original
Then, queue at the same level can use mutually idling-resource under the principle for abiding by proportional assignment.
For the cluster administrator of a Hadoop cluster, he also compares other than the isolation scheduling use for being concerned about resource
It is concerned about each user has used how many resource on earth, to formulate corresponding resource allocation plan according to user resources service condition
Slightly, or situation is actually used according to user resources and upgrade cluster, however there is presently no any technical solutions to meet this side
The demand in face.
Summary of the invention
Purpose of this disclosure is to provide a kind of Hadoop cluster user resource monitoring method and systems, for monitoring single use
The resource service condition at family.
To achieve the goals above, the disclosure provides a kind of Hadoop cluster user resource monitoring method, wherein including with
Lower step:
From the job log information of the active and standby node of Hadoop cluster acquisition user；
Data parameters from the job log information of user according to setting extract corresponding useful data；And
The useful data is counted according to statistical parameter, obtains corresponding statistical data.
Optionally, the useful data includes user data, work data and resource data.
Optionally, when the statistical parameter in the method is respectively User Identity, job identification, job run
Between, total workload, the unit time submit workload, the corresponding total run time of total workload, use stock number, total workload pair
Answer it is total using stock number, occupy stock number, the corresponding total occupancy stock number of total workload, be preempted stock number and total workload
It is corresponding to be always preempted one of stock number or a variety of；
Accordingly, the user data is user name；
The work data is the runing time of job identification and each operation；
The resource data is the quilt using stock number, the occupancy stock number of each operation and each operation of each operation
Preempting resources amount.
Optionally, the Hadoop cluster user resource monitoring method further include:
It sorts to the statistical data according to preset parameters sortnig.
Optionally, the Hadoop cluster user resource monitoring method further include:
Inquiry instruction is received, according to the querying condition in inquiry instruction, inquiry meets inquiry from obtained statistical data
The statistical data of condition.
Optionally, the Hadoop cluster user resource monitoring method further include: when receiving inquiry instruction, from institute
It states and parses the statistical parameter in the querying condition in inquiry instruction.
Optionally, the Hadoop cluster user resource monitoring method further include: output meets the statistics of querying condition
Data.
To achieve the goals above, the disclosure additionally provides a kind of Hadoop cluster user resource monitoring, comprising:
Acquisition module acquires the job log information of user for the active and standby node from Hadoop cluster in real time；
Data extraction module extracts corresponding for the data parameters from the job log information of user according to setting
Useful data；With
Statistical module obtains corresponding statistical data for counting the useful data according to statistical parameter.
Optionally, the acquisition module includes:
Interface call unit, for calling Yarn API, by the Yarn API from the active and standby node of Hadoop cluster
Collect the job log information of user；With
Data saving unit, for the job log information for the user being collected into be stored in the database of Data Persistence Layer.
Optionally, the acquisition module further include:
Format conversion unit, for converting the operation day for the user that the Yarn API is collected according to predetermined format rule
The format of will information；Accordingly, the job log information of the user after format transformation is stored in data by the data saving unit
The database of persistent layer.
Optionally, the Hadoop cluster user resource monitoring further includes interactive interface, is referred to for input inquiry
Enable and export the statistical data for meeting querying condition.
Optionally, the statistical module receives inquiry instruction, and institute is parsed from the querying condition in the inquiry instruction
Statistical parameter is stated, and the statistical data is returned into the interactive interface.
Optionally, the Hadoop cluster user resource monitoring further includes enquiry module, interacts boundary with described
Face is connected, and receives inquiry instruction, and according to the querying condition in inquiry instruction, inquiry meets inquiry from obtained statistical data
The statistical data of condition, and the statistical data inquired is returned to the interactive interface.
Optionally, the interactive interface is web application interface.
Through the above technical solutions, resource allocation management mode of the disclosure based on Yarn, is acquired by Yarn API and is used
The job log information at family collects operation start-stop runing time, user information, using information such as stock numbers, counts each user
Resource service condition.The disclosure monitors each user in cluster in real time while not influencing Hadoop group system performance
Resource service condition enables cluster administrator accurately to grasp each user to the service condition of resource, and according to user to money
The service condition in source more reasonably distributes available resource for user, so as to more reasonably adjust cluster resource,
Dilatation whether is needed for cluster and how dilatation provides foundation.
Other feature and advantage of the disclosure will the following detailed description will be given in the detailed implementation section.
Detailed description of the invention
Attached drawing is and to constitute part of specification for providing further understanding of the disclosure, with following tool
Body embodiment is used to explain the disclosure together, but does not constitute the limitation to the disclosure.In the accompanying drawings:
Fig. 1 is the Hadoop cluster user resource monitoring method flow chart that an embodiment of the present disclosure provides；
Fig. 2 is the functional block diagram for the Hadoop cluster user resource monitoring that an embodiment of the present disclosure provides；
Fig. 3 is the integrated stand composition for the Hadoop cluster user resource monitoring that an embodiment of the present disclosure provides；
Fig. 4 is the disclosure one functional structure chart for implementing the acquisition module provided；
Fig. 5 is the disclosure one functional structure chart for implementing the statistical module provided；
Fig. 6 is the disclosure one functional structure chart for implementing another statistical module provided.
Description of symbols
1,300-2-data extraction module of acquisition module
310-interface call units
320-data saving units
330-format conversion units
3,200-4-acquisition database of statistical module
210a, 210b-data extracting unit
220a, 220b-statistic unit
230a-query unit
5-result databases
100-Web 400-big data platforms of applying unit
Specific embodiment
It is described in detail below in conjunction with specific embodiment of the attached drawing to the disclosure.It should be understood that this place is retouched
The specific embodiment stated is only used for describing and explaining the disclosure, is not limited to the disclosure.
In the disclosure, in the absence of explanation to the contrary, the Hadoop cluster is one using Yarn (Yet
Another Resource Negotiator) mode manages the big data platform of cluster resource.
User data refers to data relevant to Hadoop cluster user identity, such as user name.
Data relevant to the operation that Hadoop cluster user is submitted that work data refers to, such as: each operation of job identification
Runing time etc..
Resource data refers to data relevant to resource, such as: (i.e. the operation uses the use stock number of each operation in total
Stock number), the occupancy stock number of each operation (resource in addition to distributing to the operation, in the case where other resources are idle
Used beyond distribution stock number resource), each operation be preempted stock number (be originally used for the operation distribution resource, but
It is the resource used by other operations) etc..
Data parameters refer to the foundation parameter when acquiring data, such as User Identity (ID), job identification (ID), fortune
Row time etc..
Useful data refers to the data arrived according to data parameter acquisition, i.e., the data that user is concerned about in the disclosure, for system
The initial data of meter.
Statistical parameter refers to the foundation parameter for using when statistical data, such as: User Identity, job identification, work
Industry runing time, total workload, unit time submit workload, the corresponding total run time of total workload, use stock number, total
Workload it is corresponding it is total using stock number, occupy stock number, the corresponding total occupancy stock number of total workload, be preempted stock number and
Always workload is corresponding is always preempted stock number etc..
As shown in Figure 1, for the Hadoop cluster user resource monitoring method flow chart that one embodiment of the disclosure provides, it is described
Method includes the following steps:
Step S1, from the job log information of the active and standby node of Hadoop cluster acquisition user.For example, passing through calling
The mode of Yarn API collects the job log information that user submits from the active and standby node of Hadoop cluster.Such as pass through
Restful api interface calls the details http for checking appointed task to send link GET http: // 192.168.1.1:
9009/ws/v1/cluster/app s/ job number, then can return to the detailed log information of the operation.
For example, the job log information for being collected into user submission is as follows:
ID:application_1531122299684_0056
User:aisinobi
Name:select substr(b.flbm,1,5),sum(je)zj...desc(Stage-4)
Application Type:MAPREDUCE
Queue:ysbbyjkfb
Application Tags:
State:FINISHED
FinalStatus:SUCCEEDED
Started:Mon Jul 09 19:02:54+0800 2018
Elapsed:22sec
Tracking URL:History
Total Resource Preempted:<memory:0,vCores:0>
Total Number of Non-AM Containers Preempted:0
Total Number of AM Containers Preempted:0
Resource Preempted from Current Attempt:<memory:0,vCores:0>
Number of Non-AM Containers Preempted from Current Attempt:0
Aggregate Resource Allocation:79536MB-seconds,44vcore-seconds
Step S2, the data parameters from the job log information of user according to setting extract corresponding useful data.
For example, it is entitled that the useful data therefrom extracted is respectively as follows: user using " ID ", " User ", " Name " etc. as data parameters
" aisinobi ", job identification (ID) are " application_1531122299684_0056 ", " robbing in resource data project
Accounting for resource (Total Resource Preempted) " corresponding resource data is 0；" it is preempted resource (Resource
Preempted from Current Attempt) " corresponding resource data is 0；" use resource (Aggregate
Resource Allocation) " corresponding resource data is 79536MB-seconds, 44vcore-seconds etc..
Step S3 counts the useful data according to statistical parameter, obtains corresponding statistical data.
By the job log information of all users in the available Hadoop cluster of step S1, wherein each user can
One or more operations can be submitted.On the basis of the various useful datas that step S3 is extracted in step s 2, joined according to statistics
It is several that the various useful datas are inquired, are summarized.For example, can be inquired with " User Identity " for statistical parameter
The job log information of all same subscriber identity.Cooperate statistical parameter " job identification ", can inquire to obtain all
Operation.Cooperate statistical parameter " total workload ", the number of jobs inquired is added, then the available user submits total
Workload.If cooperating statistical parameter " using stock number ", each of the user can be inquired according to " using stock number "
The use stock number of each operation is added summation by the use stock number of a operation, and the available user submits total
Workload it is corresponding it is total use stock number.
Wherein, in order to increase the convenience for understanding user resources service condition, the disclosure is using an interactive interface come defeated
Enter querying condition and inquiry instruction, and the statistical data inquired is shown in interactive interface.By being set to interactive interface
Meter, can need to only select different query terms with preset various query terms, user, can be combined into different querying conditions.It is right
In inquire come statistical data output, can preset a variety of way of outputs, such as form, graphic form, motion graphics
Formula, even speech form.
As a specific embodiment, the step S3 can be specifically embodied as following steps:
Step S31a counts the useful data according to statistical parameter, obtains corresponding statistical data, and be stored in
In result database.For example, respectively according to parametric statistics " User Identity ", " total workload ", " unit time submission operation
Amount ", " using stock number ", " total workload corresponding total use stock number ", " accounts at " total workload corresponding total run time "
With stock number ", " total workload corresponding total occupancy stock number ", " being preempted stock number " and " total workload is corresponding always to be robbed
Account for stock number " the corresponding statistical data of each user is obtained, and store into result database.
Step S32a, when receiving from the inquiry instruction of interactive interface, according to the querying condition in inquiry instruction,
Inquiry meets the statistical data of querying condition in the result database, and returns to the interactive interface.
The present embodiment corresponds to the situation of big data quantity, in order to quickly return to user as a result, the present embodiment is first according to setting
Fixed statistical parameter is counted, is calculated,, can be fast in user query by statistical data storage into result database
Speed provides result for user.
As another specific embodiment, the step S3 can be specifically embodied as following steps:
Step S31b, when receiving from the inquiry instruction of interactive interface, from the inquiry item in the inquiry instruction
Statistical parameter is parsed in part.For example, user generates inquiry item by the query term in combination interactive interface.In the present embodiment
In, the query term corresponds to statistical parameter, thus, statistical parameter can be parsed from querying condition, such as: " user name " is
" the total workload " of xxxx and " total workload corresponding total use stock number ".
Step S32b counts the useful data according to statistical parameter, obtains corresponding statistical data, and return to
The interactive interface.Specifically, it is xxxx by " user name ", inquires the entitled xxxx's of user from collected useful data
Use stock number in all work datas and each operation, according to statistical parameter " total workload ", the quantity of assignment statistics,
According to statistical parameter " total workload corresponding total use stock number ", the use stock number of each operation is added, is always made
Use stock number.Finally user interface is returned to using stock number by total workload of the entitled xxxx of user and always.
The present embodiment corresponds to the situation of lightweight data volume, is counted, is calculated again in user query, due to number
It is little according to amount, thus can achieve the purpose quickly returned the result to user.
It in one embodiment, further include being arranged according to preset parameters sortnig the statistical data after above mentioned step S3
Sequence.The parameters sortnig is, for example, total workload, always using stock number etc., thus can provide sorted statistical data,
Cluster administrator can be made to be easier to understand the resource service condition of collection user inside the group.
The Hadoop cluster user resource monitoring method provided by the disclosure, can obtain and individually use in Hadoop cluster
The resource situation at family more reasonably distributes money to the service condition of resource according to each user convenient for cluster administrator for user
Source, more reasonably adjustment cluster resource.
As shown in Fig. 2, the functional block diagram of the Hadoop cluster user resource monitoring provided for one embodiment of the disclosure.
The Hadoop cluster user resource monitoring includes acquisition module 1, data extraction module 2 and statistical module 3, wherein institute
Job log information of the acquisition module 1 from the active and standby node of Hadoop cluster acquisition user is stated, and by the job log information
It is stored in acquisition database 4.The data extraction module 2 in acquisition database 4 in the job log information of user from having extracted
With data, such as user data, work data and resource data, and it is sent to the statistical module 3.The statistical module 3
The corresponding user data of parametric statistics, work data or/and resource data obtain result data according to statistics, and are stored in result
Database 5.
Specifically, as shown in figure 3, being the whole of the Hadoop cluster user resource monitoring that one embodiment of the disclosure provides
Body architecture diagram.
Big data platform 400 therein is to want the Hadoop cluster of counting user resource service condition, including
Distributed computing framework Mapreduce, Tool for Data Warehouse Hive, distributed data base Hbase in Hadoop aggregated structure,
Computing engines Spark and file system Hdfs, and configure Hadoop cluster and carry out resource management using Yarn, it is distributed by Yarn
Cluster resource.
Acquisition module 300 is a lightweight module, and Yarn API is called to monitor Yarn scheduling of resource service condition in real time
Data.Under the premise of not influencing Hadoop clustering performance, Yarn API acquires the active and standby node of cluster, will be collected
Data are stored in the database of Data Persistence Layer.In order to be different from the storage location of statistical data, acquisition data will be stored with
Database is known as acquisition database.The result database of storage statistical data is similarly positioned in Data Persistence Layer.Wherein, collected
Data include: job identification (ID), user name, homework type, queue, the time started, runing time, preempting resources amount, are robbed
Account for stock number, consumed resource, queue name etc..
Specifically, the functional structure chart of the acquisition module 300 is as shown in figure 4, the acquisition module 300 includes interface tune
With unit 310 and data saving unit 320, wherein interface call unit 310 passes through the Yarn for calling Yarn API
API collects the job log information of user from the active and standby node of Hadoop cluster.Data saving unit 320 will use Yarn API
The acquisition database 4 of the job log information deposit Data Persistence Layer for the user being collected into.If collected job logging letter
The data format of data in breath is different from the data format in acquisition database 4, and the acquisition module 300 further includes that format turns
Unit 330 is changed, for converting the operation day for the user that the Yarn API is collected according to the data format in acquisition database 4
The format of will information.Then, the job log information of the user after format transformation is stored in data by the data saving unit 320
The acquisition database 4 of persistent layer.
In the present embodiment, the functional structure chart of statistical module 200 is as shown in figure 5, in the present embodiment, statistical module
200 include data extracting unit 210a, statistic unit 220a and query unit 230a.Wherein, data extracting unit 210a is from adopting
The useful datas such as user data, work data and resource data are extracted in collection database 4, send it to statistic unit
220a.Statistic unit 220a counts corresponding user data, work data and resource data according to preset statistical parameter,
And the statistical data obtained after statistics is stored in result database 5.When the user job log information in acquisition database 4
When having update, data extracting unit 210a extracts corresponding user data, work data and number of resources from the information of update
Corresponding user data, operation are counted from more new data according to preset statistical parameter again according to, statistic unit 220a
Data and resource data, and be stored in result database 5.Query unit 230a is applied by Restful interface to the Web on upper layer
Unit 100 provides data access function.Web applying unit 100 is an interactive interface, is used for input inquiry instruction, display symbol
Close the statistical data of querying condition.Query unit 230a receives the inquiry instruction that sends of Web applying unit 100, and according to looking into
Ask instruction in querying condition query result database 5, will inquire the statistical data for meeting querying condition export give Web application
Unit 100.
In another embodiment, the functional structure chart of the statistical module 200 is as shown in fig. 6, in the present embodiment, unite
Counting module 200 includes data extracting unit 210b and statistic unit 220b.Wherein, statistic unit module 200 is connect by Restful
Mouth provides data access function to the Web applying unit 100 on upper layer.Refer to when receiving the inquiry that Web applying unit 100 is sent
When enabling, data extracting unit 210b extracts the useful number such as user data, work data and resource data from acquisition database 4
According to sending it to statistic unit 220b, statistic unit 220b parses described from the querying condition in the inquiry instruction
Statistical parameter, and the data that the data extracting unit 210b is sent are counted according to statistical parameter, finally statistical data is returned
Back to the Web applying unit 100.
The concrete mode of operation is executed in the embodiment in relation to this method about the module in above-described embodiment, unit
In be described in detail, no detailed explanation will be given here.
The preferred embodiment of the disclosure is described in detail in conjunction with attached drawing above, still, the disclosure is not limited to above-mentioned reality
The detail in mode is applied, in the range of the technology design of the disclosure, a variety of letters can be carried out to the technical solution of the disclosure
Monotropic type, these simple variants belong to the protection scope of the disclosure.
It is further to note that specific technical features described in the above specific embodiments, in not lance
It in the case where shield, can be combined in any appropriate way, such as by the data extraction module and statistical module in Fig. 2
A component is merged into, is completed by a processor.In order to avoid unnecessary repetition, the disclosure is to various possible combinations
No further explanation will be given for mode.
In addition, any combination can also be carried out between a variety of different embodiments of the disclosure, as long as it is without prejudice to originally
Disclosed thought equally should be considered as disclosure disclosure of that.
Claims (14)
1. a kind of Hadoop cluster user resource monitoring method characterized by comprising
From the job log information of the active and standby node of Hadoop cluster acquisition user；
Data parameters from the job log information of user according to setting extract corresponding useful data；And
The useful data is counted according to statistical parameter, obtains corresponding statistical data.
2. Hadoop cluster user resource monitoring method according to claim 1, which is characterized in that the useful data packet
Include user data, work data and resource data.
3. Hadoop cluster user resource monitoring method according to claim 2, which is characterized in that the statistical parameter point
It Wei not User Identity, job identification, job run time, total workload, unit time submission workload, total workload pair
The total run time answered, using stock number, total workload it is corresponding it is total using stock number, to occupy stock number, total workload corresponding
Total occupancy stock number, be preempted stock number and total workload is corresponding is always preempted one of stock number or a variety of；
Accordingly, the user data is user name；
The work data is the runing time of job identification and each operation；
The resource data uses stock number, the occupancy stock number of each operation and being preempted for each operation for each operation
Stock number.
4. Hadoop cluster user resource monitoring method according to claim 1, which is characterized in that further include:
It sorts to the statistical data according to preset parameters sortnig.
5. Hadoop cluster user resource monitoring method according to claim 1, which is characterized in that further include:
Inquiry instruction is received, according to the querying condition in inquiry instruction, inquiry meets querying condition from obtained statistical data
Statistical data.
6. Hadoop cluster user resource monitoring method according to claim 1, which is characterized in that further include:
When receiving inquiry instruction, the statistical parameter is parsed from the querying condition in the inquiry instruction.
7. Hadoop cluster user resource monitoring method according to claim 5 or 6, which is characterized in that further include: output
Meet the statistical data of querying condition.
8. a kind of Hadoop cluster user resource monitoring characterized by comprising
Acquisition module acquires the job log information of user for the active and standby node from Hadoop cluster in real time；
Data extraction module extracts corresponding useful for the data parameters from the job log information of user according to setting
Data；With
Statistical module obtains corresponding statistical data for counting the useful data according to statistical parameter.
9. Hadoop cluster user resource monitoring according to claim 8, which is characterized in that the acquisition module packet
It includes:
Interface call unit is collected by the Yarn API from the active and standby node of Hadoop cluster for calling Yarn API
The job log information of user；With
Data saving unit, for the job log information for the user being collected into be stored in the database of Data Persistence Layer.
10. Hadoop cluster user resource monitoring according to claim 9, which is characterized in that the acquisition module
Further include:
Format conversion unit, for converting the job logging letter for the user that the Yarn API is collected according to predetermined format rule
The format of breath；Accordingly, the job log information of the user after format transformation is stored in lasting data by the data saving unit
The database of layer.
11. Hadoop cluster user resource monitoring according to claim 8, which is characterized in that further include interactive boundary
Face instructs for input inquiry and exports the statistical data for meeting querying condition.
12. Hadoop cluster user resource monitoring according to claim 11, which is characterized in that the statistical module
The inquiry instruction is received, parses the statistical parameter from the querying condition in the inquiry instruction, and by the statistics
Data return to the interactive interface.
13. Hadoop cluster user resource monitoring according to claim 11, which is characterized in that further include inquiry mould
Block is connected with the interactive interface, receives the inquiry instruction, according to the querying condition in inquiry instruction, from what is obtained
Inquiry meets the statistical data of querying condition in statistical data, and the statistical data inquired is returned to interaction circle
Face.
14. Hadoop cluster user resource monitoring according to claim 11, which is characterized in that the interactive interface
For web application interface.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811573182.6A CN109815008A (en) | 2018-12-21 | 2018-12-21 | Hadoop cluster user resource monitoring method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811573182.6A CN109815008A (en) | 2018-12-21 | 2018-12-21 | Hadoop cluster user resource monitoring method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109815008A true CN109815008A (en) | 2019-05-28 |
Family
ID=66602260
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811573182.6A Pending CN109815008A (en) | 2018-12-21 | 2018-12-21 | Hadoop cluster user resource monitoring method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109815008A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110609742A (en) * | 2019-09-25 | 2019-12-24 | 苏州浪潮智能科技有限公司 | Method and device for configuring queues of Kubernetes scheduler |
| CN112749055A (en) * | 2020-12-29 | 2021-05-04 | 拉卡拉支付股份有限公司 | Resource consumption metering method and device, electronic equipment and storage medium |
| WO2023246330A1 (en) * | 2022-06-20 | 2023-12-28 | 华为云计算技术有限公司 | Resource adjustment method and apparatus, computing device cluster, and readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110154339A1 (en) * | 2009-12-17 | 2011-06-23 | Electronics And Telecommunications Research Institute | Incremental mapreduce-based distributed parallel processing system and method for processing stream data |
| CN102939594A (en) * | 2010-02-22 | 2013-02-20 | 维图斯瑞姆有限公司 | Methods and apparatus related to migration of customer resources to virtual resources within a data center environment |
| CN106886527A (en) * | 2015-12-16 | 2017-06-23 | 北京奇虎科技有限公司 | The method for computing data and device of a kind of service-oriented |
| US20180219748A1 (en) * | 2017-02-01 | 2018-08-02 | International Business Machines Corporation | Monitoring a resource consumption of an application |
| CN108829509A (en) * | 2018-05-03 | 2018-11-16 | 山东汇贸电子口岸有限公司 | Distributed container cluster framework resources management method based on domestic CPU and operating system |
| CN108848157A (en) * | 2018-06-12 | 2018-11-20 | 郑州云海信息技术有限公司 | A kind of method and apparatus of Kubernetes cluster container monitors |
-
2018
- 2018-12-21 CN CN201811573182.6A patent/CN109815008A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110154339A1 (en) * | 2009-12-17 | 2011-06-23 | Electronics And Telecommunications Research Institute | Incremental mapreduce-based distributed parallel processing system and method for processing stream data |
| CN102939594A (en) * | 2010-02-22 | 2013-02-20 | 维图斯瑞姆有限公司 | Methods and apparatus related to migration of customer resources to virtual resources within a data center environment |
| CN106886527A (en) * | 2015-12-16 | 2017-06-23 | 北京奇虎科技有限公司 | The method for computing data and device of a kind of service-oriented |
| US20180219748A1 (en) * | 2017-02-01 | 2018-08-02 | International Business Machines Corporation | Monitoring a resource consumption of an application |
| CN108829509A (en) * | 2018-05-03 | 2018-11-16 | 山东汇贸电子口岸有限公司 | Distributed container cluster framework resources management method based on domestic CPU and operating system |
| CN108848157A (en) * | 2018-06-12 | 2018-11-20 | 郑州云海信息技术有限公司 | A kind of method and apparatus of Kubernetes cluster container monitors |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110609742A (en) * | 2019-09-25 | 2019-12-24 | 苏州浪潮智能科技有限公司 | Method and device for configuring queues of Kubernetes scheduler |
| CN110609742B (en) * | 2019-09-25 | 2023-01-06 | 苏州浪潮智能科技有限公司 | Method and device for configuring queues of Kubernetes scheduler |
| CN112749055A (en) * | 2020-12-29 | 2021-05-04 | 拉卡拉支付股份有限公司 | Resource consumption metering method and device, electronic equipment and storage medium |
| WO2023246330A1 (en) * | 2022-06-20 | 2023-12-28 | 华为云计算技术有限公司 | Resource adjustment method and apparatus, computing device cluster, and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110276182B (en) | API distributed current limiting realization method | |
| CN109815008A (en) | Hadoop cluster user resource monitoring method and system | |
| CN105139505B (en) | The non-timed reservation remote queue method and system that a kind of banking is handled | |
| CN111901249B (en) | Service flow limiting method, device, equipment and storage medium | |
| CN106294472A (en) | The querying method of a kind of Hadoop data base HBase and device | |
| CN108268372A (en) | Mock test processing methods, device, storage medium and computer equipment | |
| CN107395909A (en) | A kind of far management system and method based on cloud computing management platform | |
| CN109542608A (en) | A kind of cloud artificial tasks dispatching method based on mixing queuing network | |
| CN110308980A (en) | Batch processing method, device, equipment and the storage medium of data | |
| CN110264039A (en) | A kind of generation method and equipment of working report | |
| CN109257293A (en) | A kind of method for limiting speed, device and gateway server for network congestion | |
| CN108337275A (en) | Task distribution method, device and equipment for Distributor | |
| CN103503412B (en) | For the method and device of scheduling resource | |
| CN108681481A (en) | The processing method and processing device of service request | |
| Khazaei et al. | Performance analysis of cloud computing centers | |
| Buist et al. | A Java library for simulating contact centers | |
| CN111340397A (en) | Intelligent distribution method, server and storage medium for real estate channel customers | |
| CN111124640A (en) | Task allocation method and system, storage medium and electronic device | |
| CN111768174A (en) | Activity management method, apparatus, device and medium | |
| CN109002364A (en) | Optimization method, electronic device and the readable storage medium storing program for executing of interprocess communication | |
| CN103605575A (en) | Cloud Foundry platform application scheduling system and Cloud Foundry platform application scheduling method | |
| Shifrin et al. | Optimal control of VNF deployment and scheduling | |
| CN106534303B (en) | A kind of load-balancing method and device applied to reconciliation system | |
| CN106453557A (en) | Two-time scale dynamic bidding and resource management algorithm for user in IaaS service | |
| CN112749008A (en) | Cloud resource distribution system based on OpenStack and construction method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190528 |