CN104022913A - Test method and device for data cluster - Google Patents
Test method and device for data cluster Download PDFInfo
- Publication number
- CN104022913A CN104022913A CN201310697802.8A CN201310697802A CN104022913A CN 104022913 A CN104022913 A CN 104022913A CN 201310697802 A CN201310697802 A CN 201310697802A CN 104022913 A CN104022913 A CN 104022913A
- Authority
- CN
- China
- Prior art keywords
- test
- service
- service order
- existing network
- result data
- 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
Landscapes
- Debugging And Monitoring (AREA)
Abstract
The invention discloses a test method and device for data cluster, and belongs to the field of data cluster. The method comprises the following steps: obtaining a test authority which is isolated with the authority of a data cluster processing service; obtaining test computing resources according to the computing resources used by the data cluster processing service; obtaining current network operation flow and current network result data of the service; carrying out reduction processing on the current network operation flow to obtain test service flow; according to the test computing resources and the test authority, executing corresponding operation of each service instruction in the test service flow to obtain test result data; and carrying out consistency comparison on the current network result data and test result data and outputting comparison results. According to the test method and device for data cluster, the production environment of data cluster is directly utilized to carry out test without building a test scene additionally, thereby reducing human input and cost; the production environment of data cluster is subjected to full coverage, thereby improving coverage degree, furthermore, improving test quality, and reducing test missing rate.
Description
Technical field
The present invention relates to data cluster field, particularly a kind of method of testing for data cluster and device.
Background technology
Large data sets group refers to that thousands of computer connects computer cluster formed, that can jointly complete the tasks such as data storage, data calculating by network.
The data volume that large data sets group is carried is huge, and business complexity is also very high, and therefore, for normal and stable operation, large data sets group must have high stability.And in order to ensure large data sets group's high stability, need to test large data sets group's production environment.
But in relevant method of testing, be all generally the similar test scene of production environment constructing with large data sets group, under this test scene, test, build the cost that test scene expends very high; Because the parameters such as business diversity, business complexity and data volume of constructed test scene and large data sets group's production environment have very large gap, easily cause test result inaccurate, the existing defect of production environment that is not easy to find large data sets group, false pass rate is high.
Summary of the invention
In order to solve the problem of prior art, the embodiment of the present invention provides a kind of method of testing for data cluster.Described technical scheme is as follows:
First aspect, provides a kind of method of testing for data cluster, and described method comprises:
Obtain test authority, described test authority and data cluster are processed the authority isolation of business;
According to data cluster, process business computational resource used, obtain measuring and calculation resource;
Obtain described data cluster to the existing network operations flows of described business and existing network result data, described existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure;
Described existing network operations flows is reduced to processing, obtain test service stream, described test service stream comprises processes received service order in described business procedure;
Based on described measuring and calculation resource and described test authority, carry out operation corresponding to each service order in described test service stream, obtain the test result data of described test service stream, described test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of described indices;
Described existing network result data and described test result data are carried out to consistency contrast, output comparing result.
Second aspect, provides a kind of testing apparatus for data cluster, and described device comprises:
Test authority acquiring module, for obtaining test authority, described test authority and data cluster are processed the authority isolation of business;
Resource acquisition module, for processing business computational resource used according to data cluster, obtains measuring and calculation resource;
Existing network business acquisition module, for obtaining described data cluster to the existing network operations flows of described business and existing network result data, described existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure;
Recovery module, for described existing network operations flows being reduced to processing, obtains test service stream, and described test service stream comprises processes received service order in described business procedure;
Test result data acquisition module, be used for based on described measuring and calculation resource and described test authority, carry out operation corresponding to each service order in described test service stream, obtain the test result data of described test service stream, described test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of described indices;
Contrast module, for described existing network result data and described test result data are carried out to consistency contrast, output comparing result.
The beneficial effect that the technical scheme that the embodiment of the present invention provides is brought is:
The method and apparatus that the embodiment of the present invention provides, by existing network operations flows being reduced to process in data cluster, obtain processing the service order receiving in business procedure, again carry out the operation that each service order is corresponding, obtain the test result data of indices, by the real data of indices and test result data are carried out to consistency contrast, output comparing result.The production environment of the direct usage data cluster of this method of testing is tested, and without additionally building test scene, greatly reduced human input and the cost in test process, and the production environment of data cluster is carried out to full dose covering, strengthened coverage, further, improve test mass, reduced false pass rate.
Accompanying drawing explanation
In order to be illustrated more clearly in the technical scheme in the embodiment of the present invention, below the accompanying drawing of required use during embodiment is described is briefly described, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is the flow chart of a kind of method of testing for data cluster of providing of the embodiment of the present invention;
Fig. 2 a is the flow chart of a kind of method of testing for data cluster of providing of the embodiment of the present invention;
Fig. 2 b is the authority schematic diagram that the embodiment of the present invention provides;
Fig. 2 c is the schematic diagram that obtains test service stream that the embodiment of the present invention provides;
Fig. 2 d is the adjusting service order schematic diagram that the embodiment of the present invention provides;
Fig. 2 e is the output comparing result schematic diagram that the embodiment of the present invention provides;
Fig. 2 f is the high-level schematic functional block diagram that the embodiment of the present invention provides;
Fig. 2 g is the data cluster frame diagram that the embodiment of the present invention provides;
Fig. 2 h is the defect number design sketch that the embodiment of the present invention provides;
Fig. 2 i is the false pass rate design sketch that the embodiment of the present invention provides;
Fig. 3 is a kind of testing apparatus structural representation for data cluster that the embodiment of the present invention provides.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Embodiment based in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.
For convenience of explanation, will be introduced data cluster.Take this data cluster as Hadoop cluster be example, Hadoop cluster comprises HDFS(Hadoop Distributed File System, distributed file system) and mapping stipulations MapReduce system, HDFS system is for storing the file of Hadoop cluster, the data of high transmission rates access application are provided, be applicable to having the application program of super large data set, MapReduce system is used for realizing Distributed Calculation.HDFS and MapReduce system can run on large commercial machine cluster.
Fig. 1 is the flow chart of a kind of method of testing for data cluster of providing of the embodiment of the present invention, and referring to Fig. 1, described method comprises:
100, obtain test authority, this test authority and data cluster are processed the authority isolation of business;
101, according to data cluster, process business computational resource used, obtain measuring and calculation resource;
102, obtain this data cluster to the existing network operations flows of this business and existing network result data, this existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure;
103, this existing network operations flows is reduced to processing, obtain test service stream, this test service stream comprises processes received service order in this business procedure;
104, based on this measuring and calculation resource and this test authority, carry out operation corresponding to each service order in this test service stream, obtain the test result data of this test service stream, this test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of this indices;
105, this existing network result data and this test result data are carried out to consistency contrast, output comparing result.
The method that the embodiment of the present invention provides, by existing network operations flows being reduced to process in data cluster, obtain processing the service order receiving in business procedure, again carry out the operation that each service order is corresponding, obtain the test result data of indices, by the real data of indices and test result data are carried out to consistency contrast, output comparing result.The production environment of the direct usage data cluster of this method of testing is tested, and without additionally building test scene, greatly reduced human input and the cost in test process, and the production environment of data cluster is carried out to full dose covering, strengthened coverage, further, improve test mass, reduced false pass rate.
Alternatively, based on this measuring and calculation resource and this test resource, carry out operation corresponding to each service order in this test service stream, before obtaining the test result data of this indices, the method also comprises:
Regulate the service order in this test service stream, the service order based on after regulating is carried out test process.
Alternatively, regulate the service order in this test service stream, the service order based on after regulating is carried out test process and is comprised:
The concurrency that regulates the service order in this test service stream, this concurrency comprises at least one in instruction concurrency and mapping/stipulations MR concurrency;
According to the concurrency after regulating, issuing service instruction, the service order based on issuing is carried out test process.
Alternatively, according to the concurrency after regulating, issuing service instruction comprises:
Determine the priority of each service order in this existing network operations flows in this test service stream;
According to priority order from high to low, obtain successively the service order in this test service stream, until the service order number getting arrives the instruction concurrency after regulating, issue the service order getting;
And/or, according to priority order from high to low, obtain successively the service order in this test service stream, until the number of the mapping Map task that the service order getting generates in this existing network operations flows and stipulations Reduce task and value arrive the MR concurrency after regulating, issue the service order getting.
Alternatively, determine that in this test service stream, the priority of each service order in this existing network operations flows comprises:
According to the shared operation duration in this existing network operations flows of each service order in this test service stream, determine the priority of each service order in this existing network operations flows; Or,
The mapping Map task that in flowing according to this test service, each service order generates in this existing network operations flows and the number of stipulations Reduce task, determine the priority of each service order in this existing network operations flows.
Alternatively, the method also comprises:
Operating result data corresponding to each service order in this test service stream are stored in the default storehouse table of this data cluster, this default storehouse table is the different sink table of same data cluster with the storehouse table of the operating result data of this existing network service order of storage; Or,
Operating result data corresponding to each service order in this test service stream are stored under the predetermined directory of this data cluster to the different directories that the operating result data of this predetermined directory and storage existing network service order are same data cluster.
Alternatively, the method also comprises:
The test result data of indices in the data cluster of the test result data of this indices and old version is carried out to consistency contrast, the comparing result between output and old version.
Alternatively, this test authority comprises the authority of the data of read-only current processing business, the authority of the authority of the real data of read-only storage and read/write test result data.
Above-mentioned all optional technical schemes, can adopt any combination to form optional embodiment of the present invention, and this is no longer going to repeat them.
Fig. 2 a is the flow chart of a kind of method of testing for data cluster of providing of the embodiment of the present invention, and referring to Fig. 2 a, described method comprises:
200, obtain test authority, this test authority and data cluster are processed the authority isolation of business.
In embodiments of the present invention, before obtaining this measuring and calculation resource, obtain test authority.Tester can login test account, obtains test authority corresponding to this test account, thereby obtains measuring and calculation resource, based on this test authority and this measuring and calculation resource, this data cluster is tested.
Wherein, the project that the project that this test authority comprises comprises with the authority of this processing business is corresponding and isolation mutually one by one, existing network resource is damaged avoiding.In the authority of test authority and the business of processing, the authority setting of each project can be identical, also can be different.
As, referring to Fig. 2 b, this data cluster comprises a plurality of users pond, user pond comprises a plurality of user's groups, user's group comprises a plurality of users, the computational resource that each user pond is corresponding certain, and the corresponding user in this user pond can process business by the computational resource based on user pond.This authority list of processing business understands the authority of each user when the business of processing.A plurality of users pond comprising according to this data cluster, a plurality of test subscribers pond is set, and obtain this test authority, this test authority shows the operation that each test subscriber can carry out when testing, and test subscriber can carry out based on the corresponding measuring and calculation resource in test subscriber pond when test.
Wherein, the authority of this processing business comprises the authority of the data of the current processing business of read/write, and the real data that current processing business is obtained writes the database of real data or the authority in catalogue.Accordingly, this test authority comprises the authority of the data of read-only current processing business, the authority of the authority of the real data of read-only storage and read/write test result data, to prevent that test process from causing damage to the real data of this data cluster.
201, according to data cluster, process business computational resource used, obtain measuring and calculation resource.
The embodiment of the present invention is applied in the production environment of data cluster, under the scene that this data cluster is tested.This data cluster needs consumption calculations resource in the process of the business of processing, corresponding, in the process of testing, needs equally consumption calculations resource, first obtains measuring and calculation resource, to according to this measuring and calculation resource, this data cluster is tested.
Wherein, this computational resource refers in data cluster that, for the resource of traffic handing capacity is provided, as MapReduce resource, the embodiment of the present invention is not construed as limiting this.
Particularly, obtain with this data cluster and process business computational resource used computational resource one to one, as this measuring and calculation resource, it is identical that this measuring and calculation resource and this data cluster are processed business computational resource size used, so that processing business procedure and test process can carry out by the computational resource based on identical, further, be convenient to the stability of this data cluster to test.Alternatively, this measuring and calculation resource and this data cluster are processed business computational resource isolation used, have avoided processing the resource contention between business procedure and test process.
202, obtain this data cluster to the existing network operations flows of this business and existing network result data, this existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure.
Wherein, this existing network operations flows has recorded the service order that this data cluster receives in the process of the business of processing.In order to obtain the practical operation situation of data cluster, in processing business procedure, gather this existing network result data, this existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure, and preserves this existing network operations flows and this existing network result data.Further, dynamically from the platforms such as ganglia monitor supervision platform, system process, network management platform, gather the existing network actual parameter of the operating result data of existing network service order and the indices of processing business procedure, and the operating result data that collect (or statistics data of operating result data) and existing network actual parameter are kept in the catalogue or storehouse table that this indices is corresponding.
Wherein, this indices comprises business key index, system core index and result data index etc., can preset or set according to testing requirement, and the embodiment of the present invention does not limit this.
Alternatively, in this data cluster, a reserved interface that allows to obtain existing network operations flows, can obtain the existing network operations flows of this data cluster to this business by this interface.In addition, this interface can comprise the group system operation informations such as self-defined water operation instruction stream, database flowing water, log recording, and the embodiment of the present invention does not limit this.
In actual application, the data volume of the operating result data of real existing network service order is very large, be not easy to directly preserve, now can adopt the mode of statistics, each field of these operating result data is counted respectively, calculated and the statistic processes such as value, calculating mean value, obtain the statistics data of this existing network service order, the operating result data using these statistics data as existing network service order.In subsequent process, can adopt the statistical identical with the operating result data of this existing network service order, operating result data to this test service instruction are added up, obtain the statistics data of this test service instruction, the statistics data of this test service instruction are as the operating result data of test service instruction, and when consistency contrasts, can be undertaken by the operating result data of the operating result data to existing network service order and test service instruction, that is to say, when the statistics data of test service instruction and the statistics data consistent of existing network service order, think the operating result data of existing network service order and the operating result data consistent of test service instruction.
203, this existing network operations flows is reduced to processing, obtain test service stream, this test service stream comprises processes received service order in this business procedure.
Particularly, this existing network operations flows has recorded a plurality of service orders that this data cluster receives in the process of the business of processing, the form of each service order may be different, this existing network operations flows is filtered and form arrangement, form with the service order in unified this existing network operations flows, analyze the performance objective of each service order in this existing network operations flows, this performance objective is replaced with to test target, obtain test operation stream, this test operation stream is carried out to scenario reduction, obtain this test service stream, this test service stream comprises processes service order and test target corresponding to each service order receiving in business procedure, so that follow-up, each test target is carried out to the operation that each service order is corresponding.
In embodiments of the present invention, this existing network operations flows is filtered with form and arranged and include but not limited to that the spcial character that this existing network operations flows is comprised replaces with space or deletion.Wherein, this additional character is newline, carriage return character, a plurality of space and annotation etc.
Alternatively, according to the corresponding relation of predefined performance objective and test target, the performance objective of service order is replaced with to test target.The performance objective of processing business procedure is corresponding one by one with test target, the service order receiving in business procedure as processed is the data that read in the directory A of this data cluster, this performance objective is the data in directory A, corresponding test target is in the test result data with real data logic isolation, the data in the catalogue a corresponding with directory A.
Further alternatively, according to the corresponding relation between predefined test operation stream and test service scene, by the business scenario reduction of this data cluster test service scene corresponding for this test operation flows, to test under the test service scene after reduction.
For example, referring to Fig. 2 c, this Fig. 2 c is depicted as the filtration of service order and form is arranged to process.This existing network operations flows is kept in log recording or data-base recording, and a service order is wherein:
“Hadoop?jar\n
Xxx.jar\t…-
Dmapred?output.dir-
OnlineTargetDirl ... #-annotation-
--user " time,
By deletions such as the additional character in this service order " n ", " t ", space, annotations, obtain:
“Hadoop?jar
Xxx.jar-
Dmapred?output.dir-
OnlineTargetDirl-
--user”
The performance objective that analysis obtains service order is " OnlineTargetDirl ", according to this performance objective and test target mapping table, performance objective is replaced with to test target " TestTargetDir1 ", and user is replaced with to test subscriber " TestUser ", obtain the service order in test operation stream:
“Hadoop?jar
Xxx.jar-
Dmapred?output.dir-
TestTargetDir1-
--TestUser”
By in the test operation stream write into Databasce obtaining, this test operation stream is carried out to scenario reduction, obtain test service stream, by this test service stream write into Databasce.
Wherein, the mapping table of performance objective and test target is as shown in table 1, comprise five fields: the database " test_db " of the database " online_db " of traffic ID (sequence number) " query_id ", real data, performance objective " online_t ", test result data and test target " test_table ", the test target that performance objective " Onlineta " is corresponding is " T table1 ".The test operation stream record sheet that target obtains after replacing is as shown in table 2, comprises two fields: traffic ID " query_id " and test operation stream " OperateFlow ".Mapping table between test operation stream and test service scene is as shown in table 3, comprises three fields: test service scene ID " scene_id ", reduction sequence " step_id " and traffic ID " query_id ".
Table 1
| query_id | online_db | online_t | test_db | test_table |
| 01039131000100019853 | Onlinedb | Onlineta | Testdb | T?table1 |
Table 2
| query_id | OperateFlow |
| … | … |
Table 3
| scene_id | step_id | query_id |
| … | … | … |
204, the concurrency that regulates the service order in this test service stream, this concurrency comprises instruction concurrency and MR(Map/Reduce, mapping/stipulations) in concurrency at least one.
Wherein, this instruction concurrency refers on a time point, this data cluster can treatable service order number, this MR concurrency refers on a time point, this data cluster can the number of treatable Map task and the number of Reduce task, or number and the value of Map task and Reduce task.The embodiment of the present invention be take number that this MR concurrency is Map task and Reduce task and value and is described as example.
In embodiments of the present invention, this test service stream comprises processes the service order that business procedure receives, and by issuing the service order in this test service stream, and carries out operation corresponding to service order issuing, to processing business process simulation, obtain the test result data of indices.Wherein, the concurrency of the service order in this test service stream is different, and the pressure that this data cluster is subject in test process is just different, by regulating this concurrency, is the pressure of adjustable this data cluster.In the embodiment of the present invention, by regulating the concurrency of service order, simulate the suffered pressure in various degree of this data cluster, to the stability of this data cluster is tested.
Alternatively, regulate from small to large the concurrency of the service order in this test service stream, so that according to the concurrency after regulating, issuing service instruction, until the preset multiple that the force value of this data cluster is original force value.This preset multiple can be definite according to testing requirement, and the embodiment of the present invention does not limit this.
By regulating this test service to flow, regulate the pressure of this data cluster, directly apply pressure to this data cluster, in conjunction with the pressure of this data cluster reality, formation is greater than the controlled pressure of this data cluster actual pressure, to verify the stability performance of this data cluster under different pressure.
205, according to the concurrency after regulating, issuing service instruction.
In embodiments of the present invention, according to the concurrency after regulating, determine the service order at every turn issuing, and issue this definite service order, to carry out operation corresponding to service order issuing, after this executable operations completes, determine the service order next time issuing, to continue issuing service instruction, until all service orders in this test service stream have been issued.
Alternatively, this step 205 comprises step (1) and step (2):
(1) determine the priority of each service order in this existing network operations flows in this test service stream.
In embodiments of the present invention, for the stability to this data cluster is tested, when simulation receives the service order in this test service stream, can regulate execution sequence, concurrency of service order etc., obtain the test result data under different scenes, the priority of each service order in this existing network operations flows be can first determine, the service order for the treatment of that simulation is carried out, issuing service instruction according to this priority, determined.
Wherein, this step (1) can comprise following any one:
(1-1) according to the shared operation duration in this existing network operations flows of each service order in this test service stream, determine the priority of each service order in this existing network operations flows.
In embodiments of the present invention, can be according to the shared operation duration in this existing network operations flows of each service order in this test service stream, determine the priority of each service order, specifically can be defined as the shared longer service order of operation duration, priority is higher, also can be defined as the shared shorter service order of operation duration, priority is higher.
This mode of determining priority can be determined according to testing requirement, when the pressure maximum that can bear this data cluster when hope is tested, can determine the shared longer service order of operation duration, priority is higher, so that during according to definite priority issuing service instruction, this data cluster is produced to larger pressure.When hope is tested the Business Processing speed of this data cluster, can determine the shared shorter service order of operation duration, priority is higher, so that during according to definite priority issuing service instruction, this data cluster is produced to less pressure, with the Business Processing speed to this data cluster, test.
(1-2) according to Map task that in this test service stream, each service order generates in this existing network operations flows and the number of Reduce task, determine the priority of each service order in this existing network operations flows.
In embodiments of the present invention, each service order can generate a plurality of Map tasks and a plurality of Reduce task, the Map task that different service orders generates is different with the number of Reduce task, the pressure that this data cluster is produced is also different, can, according to the number of Map task and Reduce task, determine the priority of service order.Specifically can be defined as the service order that MR number is larger, priority is higher, also can be defined as the service order that MR number is less, and priority is higher, and wherein MR number is number and the value of Map task and Reduce task.
With according to the shared operation duration of each service order, determine priority class seemingly, according to MR number, determine that the mode of priority can be determined according to testing requirement, do not repeat them here.
(2) according to priority order from high to low, obtain successively the service order in this test service stream, until the service order number getting arrives the instruction concurrency after regulating, issue the service order getting.
In embodiments of the present invention, regulate the concurrency of the service order in this test service stream, after each adjusting, according to priority order from high to low, obtain successively the service order in this test service stream, when the service order number getting is when instruction concurrency after this regulates equates, stop obtaining service order, issue the service order having got.
As, this test service stream comprises 10,000 service orders, this instruction concurrency is 500, according to priority order from high to low, obtains 500 service orders and issues.
In another embodiment providing in the embodiment of the present invention, this step (2) can be replaced by step (3) or step (4):
(3) according to priority order from high to low, obtain successively the service order in this test service stream, until the number of the mapping Map task that the service order getting generates in this existing network operations flows and stipulations Reduce task and value arrive the MR concurrency after regulating, issue the service order getting.
In embodiments of the present invention, regulate the concurrency of the service order in this test service stream, after each adjusting, according to priority order from high to low, obtain successively the service order in this test service stream, while getting service order at every turn, calculate the MR number of this service order, and the MR number of the service order that gets of statistics and value, when equating with the MR concurrency after this adjusting with value of the MR of the service order getting number, or during the MR concurrency after being greater than this and regulating with value of the MR number of the service order getting, stop obtaining service order, and issue the service order having got.
(4) according to priority order from high to low, obtain successively the service order in this test service stream, until the service order number getting arrives the instruction concurrency after regulating, the MR number of the service order that gets of judgement whether arrive the MR concurrency after adjusting with value, if, issue the service order getting, if not, continuation is obtained the service order in this test service stream according to priority order from high to low, until the number of the Map task that the service order getting generates in this existing network operations flows and Reduce task and value arrive the MR concurrency after regulating, issue the service order getting.
Based on Fig. 2 d, according to the shared operation duration in this existing network operations flows of each service order in this test service stream, or the Map task that in this test service stream, each service order generates in this existing network operations flows and number and the value of Reduce task, determine the priority of each service order, according to priority order from high to low, obtain successively the service order in this test service stream, whether the number of the service order that judgement gets arrives the instruction concurrency after adjusting, if, continue the service order that judgement gets MR number whether arrive the MR concurrency after adjusting with value, during MR concurrency after arrive the regulating with value of this MR number, stop obtaining service order, issue the service order having got.
As, this test service stream comprises 10,000 service orders, this instruction concurrency is 500, this MR concurrency is 20,000,, according to priority order from high to low, obtain 500 service orders, whether the summation of the MR number of 500 service orders that judgement is obtained surpasses 20,000, if, stop obtaining service order, issue the service order having got, if not, continue to obtain service order, until the summation of the MR number of the service order getting surpasses 20,000.
It should be noted that, when the number of this MR concurrency number that is Map task and Reduce task, this MR concurrency comprises Map concurrency and Reduce concurrency, whether number and the value of the Map task of the service order that judgement gets arrive the Map concurrency after adjusting, and whether the number and the value that judge the Reduce task of the service order getting arrive the Reduce concurrency after adjusting, when whether the number of the Map of the service order getting task and value arrive the Map concurrency after adjusting, and when whether the number of Reduce task and value arrive the Reduce concurrency after adjusting, stop obtaining service order, and issue the service order having got.
206, based on this measuring and calculation resource and this test authority, operation corresponding to each service order that execution issues, obtain the test result data of this test service stream, this test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of this indices.
The embodiment of the present invention is the service order issuing in this test service stream, based on this measuring and calculation resource, operation corresponding to each service order that execution issues, this data cluster is processed to the process simulation of business, obtain operating result data that in test service stream, each service order is corresponding and the test result parameter of this indices.Because this test process is to carry out based on this test authority, therefore when carrying out operation corresponding to each service order, only carry out with test authority and operating accordingly, avoid the data in data cluster to damage.
In embodiments of the present invention, operating result data that need in collecting test Business Stream, each service order is corresponding in test process and the test result parameter of this indices.Particularly, the test result parameter of acquisition operations result data and indices from the platforms such as ganglia monitor supervision platform, system process, network management platform dynamically, and the test result parameter of these operating result data and indices is kept in corresponding catalogue or storehouse table.
Alternatively, operating result data corresponding to each service order in this test service stream are stored in the default storehouse table of this data cluster, this default storehouse table is the different sink table of same data cluster with the storehouse table of the operating result data of this existing network service order of storage; Or, operating result data corresponding to each service order in this test service stream are stored under the predetermined directory of this data cluster to the different directories that the operating result data of this predetermined directory and storage existing network service order are same data cluster.
In embodiments of the present invention, the existing network result data of this data cluster is distributed in each catalogue or each storehouse table of whole cluster, and for the ease of test result data is managed, store the corresponding operating result data and catalogue and the Ku Biao logic isolation of storing the operating result data of this existing network service order of each service order in this test service stream, guaranteed the fail safe of operating result data of the existing network service order of data cluster, convenient while making existing network result data and test result data carry out consistency contrast, be convenient to the follow-up cleaning of carrying out production environment.
207, this existing network result data and this test result data are carried out to consistency contrast, output comparing result.
Particularly, the computational resource that the output of this existing network result data and this test result data is recorded to number, consumption carries out consistency contrast, output comparing result.In addition, just this existing network result data and this test result data are carried out consistency contrast, can be to this existing network result data and this test result data are sampled, and the existing network sample data and the test sample book data that extract are contrasted, or contrast again after adding up.As the existing network sample data for value type and test sample book data, respectively calculating mean value, contrasts the mean value obtaining, or for existing network sample data and the test sample book data of character type, calculate respectively non-repetitive character number, the character number obtaining is contrasted.
In addition, when this data cluster has been carried out to renewal, need to test the data cluster of redaction, when obtaining this test result data, the test result data of the data cluster of the test result data of current version and old version is carried out to consistency contrast, can better to the stability of the functional module of this renewal, test.Accordingly, the method also comprises: the test result data in the data cluster of this test result data and old version is carried out to consistency contrast, the comparing result between output and old version.
Further, can set the threshold value of indices, when a certain the existing network actual parameter of index and the difference of test result parameter have exceeded the threshold value of setting, consistency contrast is not passed through.
Wherein, for quantifiable index, can directly the test result data of test result data and old version be carried out to consistency contrast, as GC number of full, average RPC(Remote Procedure Call, remote call program) duration etc., and for not quantifiable index, can carry out curve fitting to the test result data of the test result data of current version and old version, to on two kinds of curves that obtain, carry out a little consistency contrast, relatively on two kinds of curves side-play amount a little, when this side-play amount has surpassed predetermined threshold value, consistency contrast is not passed through.
In the embodiment of the present invention, when output comparing result, can export whole comparing results, can be when consistency contrast not be passed through yet, just output-consistence contrasts the comparing result of unsanctioned index, and the embodiment of the present invention does not limit this.Wherein, the mode of output comparing result can be for index be with color alarm, the tinkle of bells alarm or mail alarm etc., and the embodiment of the present invention does not limit this yet.
Referring to Fig. 2 e, testing apparatus is from ganglia monitor supervision platform, JVM(Java Virtual Machine, Java Virtual Machine), gather existing network result data in the platform such as system process, network management platform, and be kept in database.By existing network result data and test result data record number, resource consumption contrasts, the line sampling of going forward side by side, contrasts the sample data of extraction, and sample data is added up, and statistics is contrasted.Quantifiable indicator in the test result data of the test result data of current version and old version is contrasted, to curve degree and the resource utilization of quantifiable indicator do not contrast.While having contrasted, output difference surpasses the comparing result of the index of threshold value, can represent Output rusults or mail alarm on monitoring foreground.
After this step 207, the method can also comprise: when test completes, can clear up the test result data of this indices, and reclaim test authority.
Referring to Fig. 2 f, the embodiment of the present invention can be carried out by safe handling module, Service recovery module, test module, results analyses module and these five functional modules of environment cleaning module.Wherein, safe handling module is obtained and the test authority of processing the authority isolation of business for rights of using isolation, and use result data isolation that this existing network result data and test result data are carried out to logic isolation, guarantee that the existing network result data of this data cluster can be not destroyed.Service recovery module, for existing network operations flows is carried out to Operations Analyst, is extracted useful operations flows information, and is carried out suitably Safe Transformation and scenario reduction, obtains being directly used in the test service stream of test.Test module, for by regulating this test service to flow, regulates the pressure of data cluster, under the pressure after adjusting, carries out the corresponding operation of service order issuing, and simulates the process that this data cluster is processed business, obtains this test result data.Results analyses module is for carrying out consistency contrast by this existing network result data and this test result data, output comparing result, and the test result data of the test result data of current version and old version is carried out to consistency contrast, the comparing result between output and old version.Environment cleaning module, for when test completes, is cleared up unwanted test result data, and reclaims test authority.The method that the embodiment of the present invention provides can be based on shown in Fig. 2 g frame diagram.
Software or system that the method that the embodiment of the present invention provides is applicable to large data, large cluster and under test environment, cannot assesses large cluster real quality situation, adopt technical scheme of the present invention, can break through the restriction of testing the factors such as diversity, business complexity, data volume, cluster resource in large data sets group system testing, find up hill and dale the defect of large data sets group system, assess large data sets group's capacity, filled up the blank of large data sets group method of testing.
The method that the embodiment of the present invention provides, by existing network operations flows being reduced to process in data cluster, obtain processing the service order receiving in business procedure, again carry out the operation that each service order is corresponding, obtain the test result data of indices, by the real data of indices and test result data are carried out to consistency contrast, output comparing result.The production environment of the direct usage data cluster of this method of testing is tested, and without additionally building test scene, greatly reduced human input and the cost in test process, and the production environment of data cluster is carried out to full dose covering, strengthened coverage, improve test mass, reduced false pass rate.Further, regulate the pressure of this data cluster, the stability to this data cluster under large pressure is tested, and is convenient to find the defect of this data cluster under large pressure, has further reduced false pass rate.
For the beneficial effect of the bright embodiment of the present invention more particularly, provide following test result.Fig. 2 h is the defect number design sketch that the embodiment of the present invention provides.By the method for testing that the embodiment of the present invention is provided, implement, obtain comparing result, according to comparing result, find the existing defect of this data cluster, as shown in Fig. 2 h.Through conventionally test methods such as static scanning, craft, automation, special project, company's tune, finding after the defect of this data cluster some that the method that the application embodiment of the present invention provides still can be found a large amount of defects to greatly reduce false pass rate.
Fig. 2 i is the false pass rate design sketch that the embodiment of the present invention provides, and by the method for testing that the embodiment of the present invention is provided, implements, and finds defect and evaluation problem, and calculates false pass rate.Can find out, from year October in December, 2012 to 2013, false pass rate is reduced to 0.76% from 13.30%, is reduced to 2/35ths of former false pass rate, greatly reduces false pass rate.
Fig. 3 is a kind of testing apparatus structural representation for data cluster that the embodiment of the present invention provides, referring to Fig. 3, this device comprises: test authority acquiring module 300, resource acquisition module 301, existing network business acquisition module 302, recovery module 303, test result data acquisition module 304, contrast module 305
Wherein, test authority acquiring module 300 is for obtaining test authority, and this test authority and data cluster are processed the authority isolation of business; Resource acquisition module 301 is connected for process business computational resource used according to data cluster with test authority acquiring module 300, obtains measuring and calculation resource; Existing network business acquisition module 302 is connected with resource acquisition module 301, for obtaining this data cluster to the existing network operations flows of this business and existing network result data, this existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure; Recovery module 303 is connected with existing network business acquisition module 302, for this existing network operations flows being reduced to processing, obtains test service stream, and this test service stream comprises processes received service order in this business procedure; Test result data acquisition module 304 is connected with recovery module 303, for testing authority based on this measuring and calculation resource and this, carry out operation corresponding to each service order in this test service stream, obtain the test result data of this test service stream, this test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of this indices; Contrast module 305 is connected with test result data acquisition module 304, for this existing network result data and this test result data are carried out to consistency contrast, and output comparing result.
Alternatively, this device also comprises:
Adjustment module, for regulating the service order of this test service stream, the service order based on after regulating is carried out test process.
Alternatively, this adjustment module comprises:
Concurrency regulon, for regulating the concurrency of the service order of this test service stream, this concurrency comprises at least one in instruction concurrency and mapping/stipulations MR concurrency;
Instruction issues unit, for according to the concurrency after regulating, issuing service instruction;
Test execution unit, carries out test process for the service order based on issuing.
Alternatively, this instruction issues unit and comprises:
Priority is determined subelement, for determining the priority of this each service order of test service stream in this existing network operations flows;
Instruction issues subelement, for according to priority order from high to low, obtains successively the service order in this test service stream, until the service order number getting arrives the instruction concurrency after regulating, issues the service order getting;
And/or, instruction issues subelement, also for according to priority order from high to low, obtain successively the service order in this test service stream, until the number of the mapping Map task that the service order getting generates in this existing network operations flows and stipulations Reduce task and value arrive the MR concurrency after regulating, issue the service order getting.
Alternatively, this priority determines that subelement, for according to each service order of this test service stream shared operation duration in this existing network operations flows, determines the priority of each service order in this existing network operations flows; Or,
This priority determines that subelement, for the mapping Map task that generates in this existing network operations flows according to each service order of this test service stream and the number of stipulations Reduce task, determines the priority of each service order in this existing network operations flows.
Alternatively, this device also comprises:
The first memory module, for operating result data corresponding to this test service each service order of stream being stored in the default storehouse table of this data cluster, this default storehouse table is the different sink table of same data cluster with the storehouse table of the operating result data of this existing network service order of storage; Or,
The second memory module, for operating result data corresponding to this test service each service order of stream are stored under the predetermined directory of this data cluster, the different directories that the operating result data of this predetermined directory and storage existing network service order are same data cluster.
Alternatively, this device also comprises:
Old version contrast module, for the test result data of the data cluster indices of the test result data of this indices and old version is carried out to consistency contrast, the comparing result between output and old version.
Alternatively, this test authority comprises the authority of the data of read-only current processing business, the authority of the authority of the real data of read-only storage and read/write test result data.
The device that the embodiment of the present invention provides, by existing network operations flows being reduced to process in data cluster, obtain processing the service order receiving in business procedure, again carry out the operation that each service order is corresponding, obtain the test result data of indices, by the real data of indices and test result data are carried out to consistency contrast, output comparing result.The production environment of the direct usage data cluster of this method of testing is tested, and without additionally building test scene, greatly reduced human input and the cost in test process, and the production environment of data cluster is carried out to full dose covering, strengthened coverage, further, improve test mass, reduced false pass rate.
It should be noted that: the testing apparatus for data cluster that above-described embodiment provides is when the test for data cluster, only the division with above-mentioned each functional module is illustrated, in practical application, can above-mentioned functions be distributed and by different functional modules, completed as required, the internal structure of the equipment of being about to is divided into different functional modules, to complete all or part of function described above.In addition, the testing apparatus for data cluster that above-described embodiment provides belongs to same design with the method for testing embodiment for data cluster, and its specific implementation process refers to embodiment of the method, repeats no more here.
One of ordinary skill in the art will appreciate that all or part of step that realizes above-described embodiment can complete by hardware, also can come the hardware that instruction is relevant to complete by program, described program can be stored in a kind of computer-readable recording medium, the above-mentioned storage medium of mentioning can be read-only memory, disk or CD etc.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.
Claims (16)
1. for a method of testing for data cluster, it is characterized in that, described method comprises:
Obtain test authority, described test authority and data cluster are processed the authority isolation of business;
According to data cluster, process business computational resource used, obtain measuring and calculation resource;
Obtain described data cluster to the existing network operations flows of described business and existing network result data, described existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure;
Described existing network operations flows is reduced to processing, obtain test service stream, described test service stream comprises processes received service order in described business procedure;
Based on described measuring and calculation resource and described test authority, carry out operation corresponding to each service order in described test service stream, obtain the test result data of described test service stream, described test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of described indices;
Described existing network result data and described test result data are carried out to consistency contrast, output comparing result.
2. method according to claim 1, it is characterized in that, based on described measuring and calculation resource and described test authority, carry out operation corresponding to each service order in described test service stream, before obtaining the test result data of described indices, described method also comprises:
Regulate the service order in described test service stream, the service order based on after regulating is carried out test process.
3. method according to claim 2, is characterized in that, regulates the service order in described test service stream, and the service order based on after regulating is carried out test process and comprised:
The concurrency that regulates the service order in described test service stream, described concurrency comprises at least one in instruction concurrency and mapping/stipulations MR concurrency;
According to the concurrency after regulating, issuing service instruction;
Service order based on issuing is carried out test process.
4. method according to claim 3, is characterized in that, according to the concurrency after regulating, issuing service instruction comprises:
Determine the priority of each service order in described existing network operations flows in described test service stream;
According to priority order from high to low, obtain successively the service order in described test service stream, until the service order number getting arrives the instruction concurrency after regulating, issue the service order getting;
And/or, according to priority order from high to low, obtain successively the service order in described test service stream, until the number of the mapping Map task that the service order getting generates in described existing network operations flows and stipulations Reduce task and value arrive the MR concurrency after regulating, issue the service order getting.
5. method according to claim 4, is characterized in that, determines that in described test service stream, the priority of each service order in described existing network operations flows comprises:
According to the shared operation duration in described existing network operations flows of each service order in described test service stream, determine the priority of each service order in described existing network operations flows; Or,
The mapping Map task that in flowing according to described test service, each service order generates in described existing network operations flows and the number of stipulations Reduce task, determine the priority of each service order in described existing network operations flows.
6. method according to claim 1, is characterized in that, described method also comprises:
Operating result data corresponding to each service order in described test service stream are stored in the default storehouse table of described data cluster, described default storehouse table is the different sink table of same data cluster with the storehouse table of the operating result data of the described existing network service order of storage; Or,
Operating result data corresponding to each service order in described test service stream are stored under the predetermined directory of described data cluster to the different directories that the operating result data of described predetermined directory and storage existing network service order are same data cluster.
7. method according to claim 1, is characterized in that, described method also comprises:
The test result data of indices in the data cluster of the test result data of described indices and old version is carried out to consistency contrast, the comparing result between output and old version.
8. method according to claim 1, is characterized in that, described test authority comprises the authority of the data of read-only current processing business, the authority of the authority of the real data of read-only storage and read/write test result data.
9. for a testing apparatus for data cluster, it is characterized in that, described device comprises:
Test authority acquiring module, for obtaining test authority, described test authority and data cluster are processed the authority isolation of business;
Resource acquisition module, for processing business computational resource used according to data cluster, obtains measuring and calculation resource;
Existing network business acquisition module, for obtaining described data cluster to the existing network operations flows of described business and existing network result data, described existing network result data comprises the operating result data of existing network service order and processes the existing network actual parameter of the indices of business procedure;
Recovery module, for described existing network operations flows being reduced to processing, obtains test service stream, and described test service stream comprises processes received service order in described business procedure;
Test result data acquisition module, be used for based on described measuring and calculation resource and described test authority, carry out operation corresponding to each service order in described test service stream, obtain the test result data of described test service stream, described test result data comprises operating result data that in test service stream, each service order is corresponding and the test result parameter of described indices;
Contrast module, for described existing network result data and described test result data are carried out to consistency contrast, output comparing result.
10. device according to claim 9, is characterized in that, described device also comprises:
Adjustment module, for regulating the service order of described test service stream, the service order based on after regulating is carried out test process.
11. devices according to claim 10, is characterized in that, described adjustment module comprises:
Concurrency regulon, for regulating the concurrency of the service order of described test service stream, described concurrency comprises at least one in instruction concurrency and mapping/stipulations MR concurrency;
Instruction issues unit, for according to the concurrency after regulating, issuing service instruction;
Test execution unit, carries out test process for the service order based on issuing.
12. devices according to claim 11, is characterized in that, described instruction issues unit and comprises:
Priority is determined subelement, for determining the priority of described each service order of test service stream in described existing network operations flows;
Instruction issues subelement, for according to priority order from high to low, obtains successively the service order in described test service stream, until the service order number getting arrives the instruction concurrency after regulating, issues the service order getting;
And/or, instruction issues subelement, also for according to priority order from high to low, obtain successively the service order in described test service stream, until the number of the mapping Map task that the service order getting generates in described existing network operations flows and stipulations Reduce task and value arrive the MR concurrency after regulating, issue the service order getting.
13. devices according to claim 12, it is characterized in that, described priority determines that subelement, for according to each service order of described test service stream shared operation duration in described existing network operations flows, determines the priority of each service order in described existing network operations flows; Or,
Described priority determines that subelement, for the mapping Map task that generates in described existing network operations flows according to each service order of described test service stream and the number of stipulations Reduce task, determines the priority of each service order in described existing network operations flows.
14. devices according to claim 9, is characterized in that, described device also comprises:
The first memory module, for operating result data corresponding to described test service each service order of stream being stored in the default storehouse table of described data cluster, described default storehouse table is the different sink table of same data cluster with the storehouse table of the operating result data of the described existing network service order of storage; Or,
The second memory module, for operating result data corresponding to described test service each service order of stream are stored under the predetermined directory of described data cluster, the different directories that the operating result data of described predetermined directory and storage existing network service order are same data cluster.
15. devices according to claim 9, is characterized in that, described device also comprises:
Old version contrast module, for the test result data of the data cluster indices of the test result data of described indices and old version is carried out to consistency contrast, the comparing result between output and old version.
16. devices according to claim 9, is characterized in that, described test authority comprises the authority of the data of read-only current processing business, the authority of the authority of the real data of read-only storage and read/write test result data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310697802.8A CN104022913B (en) | 2013-12-18 | 2013-12-18 | For method of testing and the device of data cluster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310697802.8A CN104022913B (en) | 2013-12-18 | 2013-12-18 | For method of testing and the device of data cluster |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104022913A true CN104022913A (en) | 2014-09-03 |
| CN104022913B CN104022913B (en) | 2015-09-09 |
Family
ID=51439495
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310697802.8A Active CN104022913B (en) | 2013-12-18 | 2013-12-18 | For method of testing and the device of data cluster |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104022913B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104539614A (en) * | 2014-12-27 | 2015-04-22 | 北京奇虎科技有限公司 | Method and device for website detection concurrency amount control |
| CN105243168A (en) * | 2015-11-11 | 2016-01-13 | 中国建设银行股份有限公司 | Data migration method and system |
| CN105577470A (en) * | 2015-12-12 | 2016-05-11 | 天津南大通用数据技术股份有限公司 | Cluster abnormity orthogonal test method based on distributed resource control |
| CN106453549A (en) * | 2016-10-08 | 2017-02-22 | 中国工商银行股份有限公司 | Distributed software version release method and system thereof |
| WO2017140084A1 (en) * | 2016-02-15 | 2017-08-24 | 中兴通讯股份有限公司 | Test method and device |
| CN109976997A (en) * | 2017-12-28 | 2019-07-05 | 北京京东尚科信息技术有限公司 | Test method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090199047A1 (en) * | 2008-01-31 | 2009-08-06 | Yahoo! Inc. | Executing software performance test jobs in a clustered system |
| US8321454B2 (en) * | 2009-09-14 | 2012-11-27 | Myspace Llc | Double map reduce distributed computing framework |
| CN103176903A (en) * | 2013-03-12 | 2013-06-26 | 百度在线网络技术（北京）有限公司 | Testing method and testing device of MapReduce distributed system program |
| CN103338132A (en) * | 2013-06-21 | 2013-10-02 | 安科智慧城市技术（中国）有限公司 | Method and system for verifying access of equipment in management platform |
-
2013
- 2013-12-18 CN CN201310697802.8A patent/CN104022913B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090199047A1 (en) * | 2008-01-31 | 2009-08-06 | Yahoo! Inc. | Executing software performance test jobs in a clustered system |
| US8321454B2 (en) * | 2009-09-14 | 2012-11-27 | Myspace Llc | Double map reduce distributed computing framework |
| CN103176903A (en) * | 2013-03-12 | 2013-06-26 | 百度在线网络技术（北京）有限公司 | Testing method and testing device of MapReduce distributed system program |
| CN103338132A (en) * | 2013-06-21 | 2013-10-02 | 安科智慧城市技术（中国）有限公司 | Method and system for verifying access of equipment in management platform |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104539614A (en) * | 2014-12-27 | 2015-04-22 | 北京奇虎科技有限公司 | Method and device for website detection concurrency amount control |
| CN104539614B (en) * | 2014-12-27 | 2018-03-27 | 北京奇安信科技有限公司 | A kind of method and apparatus of website detection number of concurrent control |
| CN105243168A (en) * | 2015-11-11 | 2016-01-13 | 中国建设银行股份有限公司 | Data migration method and system |
| CN105243168B (en) * | 2015-11-11 | 2019-08-30 | 中国建设银行股份有限公司 | A kind of data migration method and system |
| CN105577470A (en) * | 2015-12-12 | 2016-05-11 | 天津南大通用数据技术股份有限公司 | Cluster abnormity orthogonal test method based on distributed resource control |
| WO2017140084A1 (en) * | 2016-02-15 | 2017-08-24 | 中兴通讯股份有限公司 | Test method and device |
| CN106453549A (en) * | 2016-10-08 | 2017-02-22 | 中国工商银行股份有限公司 | Distributed software version release method and system thereof |
| CN106453549B (en) * | 2016-10-08 | 2019-09-13 | 中国工商银行股份有限公司 | A kind of distributed software version dissemination method and system |
| CN109976997A (en) * | 2017-12-28 | 2019-07-05 | 北京京东尚科信息技术有限公司 | Test method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104022913B (en) | 2015-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11087329B2 (en) | Method and apparatus of identifying a transaction risk | |
| CN104022913B (en) | For method of testing and the device of data cluster | |
| DE112012005037B4 (en) | Manage redundant immutable files using deduplications in storage clouds | |
| CN106897322B (en) | A kind of access method and device of database and file system | |
| CN106980669B (en) | A kind of storage of data, acquisition methods and device | |
| KR101559206B1 (en) | Method of processing log data, and system operating the same | |
| DE112019002948T5 (en) | DETERMINING AN OPTIMAL STORAGE ENVIRONMENT FOR DATA RECORDS AND FOR THE MIGRATION OF DATA RECORDS | |
| US9177020B2 (en) | Gathering index statistics using sampling | |
| CN108717661B (en) | Cluster storage and analysis method for financial industry risk early warning | |
| DE112020005323T5 (en) | ELASTIC EXECUTION OF MACHINE LEARNING WORKLOADS USING APPLICATION-BASED PROFILING | |
| CN104750826B (en) | A kind of structural data resource metadata is screened automatically and dynamic registration method | |
| CN104615785A (en) | Data storage method and device based on TYKY cNosql | |
| CN111181800A (en) | Test data processing method and device, electronic equipment and storage medium | |
| DE112021005364T5 (en) | DEFENSE TARGETED DATABASE ATTACKS THROUGH DYNAMIC HONEYPOT DATABASE RESPONSE GENERATION | |
| US11321318B2 (en) | Dynamic access paths | |
| US9830383B2 (en) | Decision table decomposition using semantic relations | |
| CN108306780B (en) | Cloud environment-based virtual machine communication quality self-optimization system and method | |
| CN109359205A (en) | A kind of remote sensing image cutting method and equipment based on geographical grid | |
| CN109656945A (en) | A kind of information processing method | |
| CN206421382U (en) | A kind of data handling system | |
| CN109033444A (en) | The method and device across organizational boundary's data sharing is realized based on object storage technology | |
| CN108873781A (en) | A kind of Full-automatic digital equipment | |
| CN111552956A (en) | Role authority control method and device for background management | |
| CN104598321A (en) | Intelligent big data processing method and device | |
| CN104184629A (en) | OTA hotel system-based performance monitoring method |
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 |