CN103793309B - A kind of batch service method for early warning and device - Google Patents

A kind of batch service method for early warning and device Download PDF

Info

Publication number
CN103793309B
CN103793309B CN201210423143.4A CN201210423143A CN103793309B CN 103793309 B CN103793309 B CN 103793309B CN 201210423143 A CN201210423143 A CN 201210423143A CN 103793309 B CN103793309 B CN 103793309B
Authority
CN
China
Prior art keywords
batch service
average
abnormal
batch
resource
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.)
Active
Application number
CN201210423143.4A
Other languages
Chinese (zh)
Other versions
CN103793309A (en
Inventor
吴永卫
陈航
肖爱元
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Mobile Group Zhejiang Co Ltd
Original Assignee
China Mobile Group Zhejiang Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Mobile Group Zhejiang Co Ltd filed Critical China Mobile Group Zhejiang Co Ltd
Priority to CN201210423143.4A priority Critical patent/CN103793309B/en
Publication of CN103793309A publication Critical patent/CN103793309A/en
Application granted granted Critical
Publication of CN103793309B publication Critical patent/CN103793309B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The present invention is applied to database field, there is provided a kind of batch service method for early warning and device, methods described comprise the steps:System resources consumption is detected, and index is consumed using result of detection computing system resource;Time reference line is calculated, the system resource is consumed into index contrasts with the time reference line, judges whether system operation is normal;When system operation is abnormal, by the batch service Characteristic Contrast of known batch service in the business currently run and batch service list, abnormal batch service is identified;Sent a warning message according to recognition result to alarm platform.The present invention identifies abnormal batch service by the way that the system resource consumption index of calculating and time reference line contrast are judged into whether system operation is normal, by the batch service Characteristic Contrast with known batch service, and sends corresponding warning information to alarm platform.Contribute to attendant to predict abnormal batch service rapidly and make quick correctly processing decision-making.

Description

A kind of batch service method for early warning and device
Technical field
The present invention is applied to database field, more particularly to a kind of batch service method for early warning and device.
Background technology
With in business operation support system (Business&Operation Support System, BOSS), business needs That asks is continuously increased, and the type of batch service and quantity are more and more in BOSS, and batch service frequency of abnormity also increases therewith.Batch Amount service exception refers to that batch service is not initiated in normal time section, although or being initiated in normal time section, occupancy Cross multi-system resource.Batch service can seriously consume extremely central processing unit (Central Processing Unit, CPU) and The resource of input and output (input/output, I/O) bus, causes the processing speed of each business in BOSS systems to reduce, increase The stand-by period of user.
In the prior art, the SQL of multi-system resource was taken using the means of Centralized Monitoring to find (Structured Query Language, SQL) sentence and warn DBA (DatabaseAdministrator, DBA), still, because class of business is various, DBA can not understand the model of each business in depth, cause DBA not ask immediately Topic SQL statement corresponds to specific business, i.e., can not quickly solve the problems, such as the batch service exception occurred, business processing speed Spending the problem of slow is not resolved.
The content of the invention
The embodiments of the invention provide a kind of batch service method for early warning, it is intended to which solution uses Centralized Monitoring in the prior art Technology took the SQL statement of multi-system resource to find, and the batch service that can not quickly solve to have occurred is abnormal, business processing Slow-paced problem.
The embodiment of the present invention is achieved in that a kind of batch service method for early warning, and methods described comprises the steps:
System resources consumption is detected, and index is consumed using result of detection computing system resource;
Time reference line is calculated, the system resource is consumed into index contrasts with the time reference line, judges that system operation is It is no normal;
When system operation is abnormal, by the batch of known batch service in the business currently run and batch service list Service feature contrasts, and identifies abnormal batch service;
Sent a warning message according to recognition result to alarm platform.
The embodiment of the present invention also provides a kind of batch service prior-warning device, and described device includes:
Resource probe unit, for being detected to system resources consumption, and utilize result of detection computing system resource consumption Use index;
Time reference line unit, for calculating time reference line, the system resource is consumed into index and the time reference line pair Than judging whether system operation is normal;
Batch service recognition unit, when system operation is abnormal, by the business currently run and batch service list The batch service Characteristic Contrast of known batch service, identifies abnormal batch service;And
Information alert unit, for being sent a warning message according to recognition result to alarm platform.
In embodiments of the present invention, by the way that the system resource of calculating is consumed into index and time reference line contrast judgement system fortune Whether row is normal, and abnormal batch service is identified by the batch service Characteristic Contrast with known batch service, and flat to alarm Platform sends corresponding warning information.Contribute to attendant to predict abnormal batch service rapidly and make quickly correctly handling to determine Plan.
Brief description of the drawings
Fig. 1 represents the implementation process figure of batch service method for early warning provided in an embodiment of the present invention;
Fig. 2 represents the structure chart of batch service prior-warning device provided in an embodiment of the present invention;
Fig. 3 represents the structure chart of resource probe unit provided in an embodiment of the present invention;
Fig. 4 represents the structure chart of time reference line unit provided in an embodiment of the present invention;
Fig. 5 represents the structure chart of batch service recognition unit provided in an embodiment of the present invention;
Fig. 6 represents the structure chart of information alert unit provided in an embodiment of the present invention.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and It is not used in the restriction present invention.
In inventive embodiments, system is calculated by detecting operating-system resources consumption and/or Database Systems resource consumption Resource of uniting consumes index and time reference line, and both are contrasted and judges whether system operation is normal.To running abnormal system, pass through The batch service Characteristic Contrast of the business currently run and known batch service is identified into abnormal batch service, and it is flat to alarm Platform sends corresponding warning information.
Fig. 1 shows the implementation process of batch service method for early warning provided in an embodiment of the present invention, and details are as follows:
In step S101, system resources consumption is detected, and is consumed and referred to using result of detection computing system resource Mark;
In embodiments of the present invention, system resources consumption is in operating-system resources consumption and database system resources consumption It is at least one.
In embodiments of the present invention, operating-system resources are consumed every 30 seconds and/or Database Systems resource consumption is visited Survey once.
As one embodiment of the present of invention, utilization rate pcpu, memory usage that operating-system resources consumption is CPU It is at least one in pmen, and I/O utilization rates pio.
As one embodiment of the present of invention, Database Systems resource consumption consumes including statistical information, and activity industry Business consumption.Wherein:
Statistical information consumption is average login times ps_logon per second, average logic per second reads ps_logic, flat It is physical read ps_physi per second, average number of transactions ps_trans per second, average day quality ps_redos per second, average Active session ps_activ per second, the accumulative usage amount ps_undo of average rewind journal per second, average hard parsing per second It is at least one in ps_hard, and average vernier opening number ps_cursor per second.
Live traffic consumption waits pw_us, index contention to wait pw_ic, serial contention to wait pw_ for roll-back segment contention Sq, focus block contention wait at least one in pw_hotblk, and journal file synchronization wait pw_logfile.
In embodiments of the present invention, when calculating operating-system resources consumption index p sys, it is contemplated that CPU and memory usage Importance, the proportion of CPU, internal memory, and I/O three are configured to 5: 4: 1, i.e., 50% is made up of cpu resource, and 40% by interior Resource composition is deposited, 10% by I/O resource compositions.
In step s 102, time reference line is calculated, system resource is consumed into index contrasts with time reference line, judges that system is transported Whether row is normal;
As one embodiment of the present of invention, time reference line includes operating-system resources baseline and database system resource base It is at least one in line.Wherein:
Operating-system resources baseline is CPU baselines bs_cpu_N, internal memory baseline bs_io_N, and I/O baselines bs_mem_ N.Database Systems resource baseline includes login times baseline bs_logon_N, logic reads baseline bs_logic_N, physical read base Line bs_physi_N, number of transactions baseline bs_trans_N, daily record amount baseline bs_redos_N, active session baseline bs_redos_ N, the accumulative usage amount baseline bs_undo_N of rewind journal, hard parsing baseline bs_hard_N, vernier open base line bs_ Cursor_N, roll-back segment contention wait baseline bs_us_N, index contention to wait baseline bs_ic_N, serial contention to wait baseline Bs_sq_N, focus block contention wait baseline bs_hotblk_N, and journal file synchronously to wait in baseline bs_logfile_N It is at least one.
In embodiments of the present invention, operating-system resources baseline is calculated by following step:
1st, the activity duration is segmented according to the batch service of the statistics execution time;
2nd, calculate respectively operating-system resources consumption in setting time threshold value it is daily each in the period average it is equal Value M1, and in setting time threshold value within daily each period peak average M2, and calculate average M1 and average M2 average M3.
In embodiments of the present invention, the batch service of statistics can perform to the time be divided into 4 periods, and respectively 22:00- 6:00、6:00-8:00、8:00-17:00、17:00-22:00, the time threshold for calculating operating-system resources baseline is set as 2 Week.
It is described further exemplified by calculating CPU baselines bs_cpu_N, at daily morning 12, calculates utilization rate most respectively The nearly 2 weeks average M1-N of the average and average M2-N of peak within every 4 periods of CPU days, then M1-N and M2-N are averaged Obtain bs_cpu_N.Wherein, N take 1,2,3,4, bs_cpu_1 represented 22:00-6:00 CPU baselines, bs_cpu_2 are represented 6:00-8:00 CPU baselines, bs_cpu_3 are represented 8:00-17:00 CPU baselines, bs_cpu_4 are represented 17:00-22: 00 CPU baselines.
In embodiments of the present invention, Database Systems resource baseline is calculated by following step:
1st, the activity duration is segmented according to the batch service of the statistics execution time;
2nd, Database Systems resource consumption average within daily each period in setting time threshold value is calculated respectively Average M4, and in setting time threshold value within daily each period peak average M5, and calculate average M4 and Value M5 average M6.
In embodiments of the present invention, the computational methods of Database Systems resource baseline and the calculating of operating-system resources baseline Method is identical.The calculating process of each baseline is identical with above-mentioned CPU baselines bs_cpu_N calculating process.
In step s 103, will be known in the business currently run and batch service list when system operation is abnormal The batch service Characteristic Contrast of batch service, identifies abnormal batch service;
In embodiments of the present invention, when system operation is abnormal, by the business currently run and batch service list The batch service feature of known batch service is compared the abnormal batch service of identification, when system operation is normal, then wait into Row detection next time.Batch service feature includes performing period, characteristic query sentence SQL, characteristic query sentence mark SQL_ID, And batch service title.Wherein, determine that characteristic query sentence identifies SQL_ID by characteristic query sentence SQL.
In embodiments of the present invention, the step of identifying abnormal batch service be specially:
1st, regularly data base view V session are scanned, arrange out the non-NULL that system resource consumes seniority among brothers and sisters first Resource consumes the SQL_ID of seniority among brothers and sisters first in not busy wait event.
The 2nd, if system resource consumes the batch service baseline that index is higher than current time, by this SQL_ID in batch service Scanned in list.
The 3rd, SQL_ID and the SQL_ID in batch service list that the session currently run is performed to sentence SQL is contrasted, if Matching is then that known batch service is abnormal, is probably that unknown batch service is abnormal if mismatching, it is also possible to which the abnormal traffic is not It is batch service.
In step S104, sent a warning message according to recognition result to alarm platform.
As a preferred embodiment of the present invention, it is to the step of platform sends a warning message is alerted according to recognition result It is at least one in following step:
1st, when abnormal batch service is known batch service, judge whether abnormal batch performs in regular time periods, be then Batch service property abnormality warning information is sent to platform is alerted, batch service abnormal time point is otherwise sent and performs warning information To alarm platform;
2nd, when abnormal batch service is unknown batch service, unknown service exception warning information is sent to alerting platform;
3rd, when finding new batch service, or finding the new feature of existing batch service, new batch service is added Or the new feature information of existing batch service is into batch service list;
4th, when abnormal traffic is unrelated with batch service, non-batch abnormality alarming information is sent to alerting platform, and will be different Reason condition adds document library and put on record.
In embodiments of the present invention, it is assumed that known batch service A and its director B.If batch service A is abnormal, then Further compare its uptime section, if the period does not meet, alarm " batch service A non-normal hours section performs, Batch service director B please be contact immediately to stop." if the period meets, " batch service A property abnormalities, please be contacted for alarm Investigate using batch service director B and immediately!”.
For the not batch service in batch service list, divided by the different type of system resources consumption index:Qiang Xin Increasing type (Insertion Insensible, II), strong (Modification Insensible, MI) and the strong deletion type of more retrofiting (Deletion Insensible, DI), the combined business (CombinationJob, CJ) of first three types and non-batch industry Exception (Not Batch, NB) of being engaged in totally five type.In alarm, alarm can be sent according to this five kinds of different types of keywords Information, maintain easily personnel and carry out communication identification according to type and related application personnel rapidly, and carry out respective handling.
Fig. 2 shows batch service prior-warning device structure provided in an embodiment of the present invention, for the ease of description, illustrate only The part related to the embodiment of the present invention.
21 pairs of system resources consumptions of resource probe unit (Resource Probe Model, RPM) detect, and utilize Result of detection computing system resource consumes index.
Time reference line unit (Time-based Baseline Model, TBM) 22 calculates time reference line, by system resource Consume index to contrast with time reference line, judge whether system operation is normal.
Batch service recognition unit (Batch Job Identification Model, BJIM) 23 is working as system operation not just Chang Shi, the batch service Characteristic Contrast of known batch service in the business currently run and batch service list identifies different Normal batch service.
Information alert unit (Alarm Policy Model, APM) 24 is according to the recognition result of batch service recognition unit 23 Sent a warning message to alarm platform.
The embodiment of the present invention uses three tiers application framework from master-plan, is respectively:Metadata acquisition layer, data processing Layer and application layer.Three tiers application framework, interdepend again independently of each other between layers, embody the flexibility of framework, for batch Operation provides method for early warning extremely.
Resource probe unit RPM belongs to metadata acquisition layer, is responsible for providing the main frames such as monitored host CPU, internal memory and IO Source is detected and is stored in database, and basic data is provided for data analysis layer.
Time reference line unit TBM and batch service recognition unit BJIM belong to data analysis layer, are responsible for receiving acquisition layer The metadata of collection is analyzed, and regularly the data in database are calculated, and draws time reference line, and generates data processing knot Fruit, alarm foundation is provided for application layer.
Information alert unit AP M belongs to application layer, is responsible for alarm and short message upgrading of batch jobs abnormal conditions etc., for dimension Shield personnel can track, and quickly accurately handle abnormal offer decision information.
Fig. 3 shows the structure of resource probe unit provided in an embodiment of the present invention, and details are as follows:
Resource probe unit RPM21 includes operating-system resources probe module 31 and/or Database Systems resource probe mould Block 32.Wherein:
Operating-system resources probe module 31 detects operating-system resources consumption.
The detection data storehouse system resources consumption of Database Systems resource probe module 32.
As one embodiment of the present of invention, Database Systems resource probe module 32 includes statistical information probe submodule 321 and/or live traffic probe submodule 322.Wherein:
Statistical information probe submodule 321 detects statistical information consumption.
The detected event business of live traffic probe submodule 322 consumes.
As one embodiment of the present of invention, utilization rate pcpu, memory usage that operating-system resources consumption is CPU It is at least one in pmen, and I/O utilization rates pio.
As one embodiment of the present of invention, statistical information consumption is average login times ps_logon per second, averagely Logic per second reads ps_logic, average physical read ps_physi per second, average number of transactions ps_trans per second, average every Day quality ps_redos of second, average active session ps_activ per second, the accumulative usage amount of average rewind journal per second Ps_undo, average hard parsing ps_hard per second, and average vernier per second open number ps_cursor.Live traffic disappears Consumption includes roll-back segment contention and waits pw_us, index contention to wait pw_ic, serial contention to wait pw_sq, focus block contention to wait It is at least one in pw_hotblk, and journal file synchronization wait pw_logfile.
In embodiments of the present invention, when calculating operating-system resources consumption index p sys, it is contemplated that CPU and memory usage Importance, the proportion of CPU, internal memory, and I/O three are configured to 5: 4: 1, i.e., 50% is made up of cpu resource, and 40% by interior Resource composition is deposited, 10% by I/O resource compositions.
Fig. 4 shows the structure of time reference line unit provided in an embodiment of the present invention, and details are as follows:
Time reference line unit TBM22 includes operating-system resources base line module 41 and/or Database Systems resource baseline mould Block 42.Wherein:
Operating-system resources base line module 41 calculates operating-system resources baseline.
Database Systems resource base line module 42 calculates Database Systems resource baseline.
As one embodiment of the present of invention, operating-system resources base line module 41 includes activity duration fractionation submodule 411 and the first mean value computation submodule 412.Wherein:
Activity duration splits submodule 411 and the activity duration is segmented according to the batch service of the statistics execution time.
First mean value computation submodule 412 calculates operating-system resources consumption in setting time threshold value daily every respectively The average M1 of average in the individual period, and in setting time threshold value within daily each period peak average M2, And calculate average M1 and average M2 average M3.
As one embodiment of the present of invention, Database Systems resource base line module 42 includes the second mean value computation submodule 421.Second mean value computation submodule 421 calculates Database Systems resource consumption in setting time threshold value daily each respectively The average M4 of average in period, and in setting time threshold value within daily each period peak average M5, and Calculate average M4 and average M5 average M6.
As one embodiment of the present of invention, operating-system resources baseline is CPU baselines bs_cpu_N, internal memory baseline bs_ It is at least one in io_N, and I/O baselines bs_mem_N.Database Systems resource baseline is login times baseline bs_ Logon_N, logic read baseline bs_logic_N, physical read baseline bs_physi_N, number of transactions baseline bs_trans_N, daily record amount Baseline bs_redos_N, active session baseline bs_redos_N, the accumulative usage amount baseline bs_undo_N of rewind journal, hard solution Analyse baseline bs_hard_N, vernier opens base line bs_cursor_N, roll-back segment contention waits baseline bs_us_N, index contention Baseline bs_ic_N, serial contention is waited to wait baseline bs_sq_N, focus block contention to wait baseline bs_hotblk_N, Yi Ji Will file synchronization waits at least one in baseline bs_logfile_N.
Fig. 5 shows the structure of batch service recognition unit provided in an embodiment of the present invention, and details are as follows:
Batch service recognition unit BJIM31 includes batch service list block 51 and batch service characteristic module.Batch industry Business list block 51 stores known batch service, and batch service characteristic module 52 identifies abnormal batch service.
In embodiments of the present invention, by the batch industry of known batch service in the business currently run and batch service list Business feature compares the abnormal batch service of identification.Batch service feature is looked into including execution period, characteristic query sentence SQL, feature Ask sentence mark SQL_ID, and batch service title.Wherein, determine that characteristic query sentence identifies by characteristic query sentence SQL SQL_ID。
Fig. 6 shows the structure of information alert unit provided in an embodiment of the present invention, and details are as follows:
Information alert unit AP M24 includes known batch service abnormality alarming module 61, unknown batch service abnormality alarming Module 62, newly-increased batch service alarm module 63, and non-batch service abnormality alarming module 64.
Known batch service abnormality alarming module 61 judges abnormal batch when abnormal batch service is known batch service Whether business performs in regular time periods, is to send batch service property abnormality warning information to platform is alerted, and otherwise sends and criticizes Amount service exception time point performs warning information to alerting platform.
Unknown batch service abnormality alarming module 62 sends unknown business when abnormal batch service is unknown batch service Abnormality alarming information is to alerting platform.
Newly-increased batch service alarm module 63 is when the new batch service of discovery, or finds the new spy of existing batch service During sign, the new feature information of new batch service or existing batch service is added into batch service list.
Non- batch service abnormality alarming module 64 sends non-batch abnormality alarming when abnormal traffic is unrelated with batch service Information is put on record abnormal conditions addition document library to alerting platform.
In inventive embodiments, system is calculated by detecting operating-system resources consumption and/or Database Systems resource consumption Resource of uniting consumes index and time reference line, and both are contrasted and judges whether system operation is normal.To running abnormal system, pass through The batch service Characteristic Contrast of the business currently run and known batch service is identified into abnormal batch service, and it is flat to alarm Platform sends corresponding warning information.So attendant can predict possible abnormal traffic rapidly, quickly correctly handle it Decision-making, so that core system failure caused by progressively preventing abnormal batch service.
Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should It is considered as protection scope of the present invention.

Claims (13)

1. a kind of batch service method for early warning, it is characterised in that methods described comprises the steps:
System resources consumption is detected, and index is consumed using result of detection computing system resource;
Time reference line is calculated, whether just the system resource is consumed into index contrasts with the time reference line, judge system operation Often;
When system operation is abnormal, by the batch service of known batch service in the business currently run and batch service list Characteristic Contrast, abnormal batch service is identified, including:Timing is scanned to data base view, arranges out system resource consumption Resource consumes the SQL_ID of seniority among brothers and sisters first in the busy wait event of seniority among brothers and sisters first;If system resource consumes index higher than current The batch service baseline at moment, then this SQL_ID is scanned in batch service list;The session currently run is performed into sentence SQL SQL_ID contrasts with the SQL_ID in batch service list, is that known batch service is abnormal if matching, if mismatching It is that unknown batch service is abnormal or the abnormal traffic is not batch service;
Sent a warning message according to recognition result to alarm platform.
2. the method as described in claim 1, it is characterised in that the system resources consumption consumes including operating-system resources And/or Database Systems resource consumption, the time reference line include operating-system resources baseline and/or Database Systems resource base Line.
3. method as claimed in claim 2, it is characterised in that the operating-system resources consumption is central processor CPU It is at least one in utilization rate, memory usage, and input/output bus I/O utilization rates.
4. method as claimed in claim 2, it is characterised in that the Database Systems resource consumption consumes including statistical information And/or live traffic consumption;
Statistical information consumption is average login times per second, average logic per second is read, average physical read per second, flat Number of transactions per second, average per second day quality, average active session per second, average the accumulative of rewind journal per second make Dosage, average hard parsing per second, and average vernier per second are opened at least one in number;
The live traffic consumption waits for roll-back segment contention, indexes contention wait, the wait of serial contention, focus block contention etc. Treat, and journal file is synchronously waiting at least one.
5. method as claimed in claim 2, it is characterised in that the operating-system resources baseline is calculated by following step:
The activity duration is segmented according to the batch service of the statistics execution time;
The average of operating-system resources consumption average within daily each period in setting time threshold value is calculated respectively M1, and in setting time threshold value within daily each period peak average M2, and calculate the average M1 and Value M2 average M3.
6. method as claimed in claim 2, it is characterised in that the Database Systems resource baseline passes through following step meter Calculate:
The activity duration is segmented according to the batch service of the statistics execution time;
Calculate respectively the Database Systems resource consumption in setting time threshold value within daily each period average it is equal Value M4, and in setting time threshold value within daily each period peak average M5, and calculate the average M4 and Average M5 average M6.
7. the method as described in claim 1, it is characterised in that described to be sent a warning message according to recognition result to alarm platform The step of to be at least one in following steps:
When the abnormal batch service is known batch service, judge whether the abnormal batch performs in regular time periods, be Batch service property abnormality warning information is then sent to platform is alerted, batch service abnormal time point is otherwise sent and performs alarm letter Breath extremely alarm platform;
When the abnormal batch service is unknown batch service, unknown service exception warning information is sent to alerting platform;
When finding new batch service, or finding the new feature of existing batch service, add the new batch service or The new feature information of the existing batch service of person is into batch service list;
When the abnormal traffic is unrelated with batch service, send non-batch abnormality alarming information to alerting platform, and will described in Abnormal conditions add document library and put on record.
8. a kind of batch service prior-warning device, it is characterised in that described device includes:
Resource probe unit, for being detected to system resources consumption, and consumed and referred to using result of detection computing system resource Mark;
Time reference line unit, for calculating time reference line, the system resource is consumed into index and contrasted with the time reference line, is sentenced Whether disconnected system operation is normal;
Batch service recognition unit, for when system operation is abnormal by the business currently run and batch service list Know the batch service Characteristic Contrast of batch service, identify abnormal batch service, including:Timing is swept to data base view Retouch, arrange out system resource and consume the SQL_ID that resource in the busy wait event for rank first consumes seniority among brothers and sisters first;If it is Resource of uniting consumes the batch service baseline that index is higher than current time, then scans this SQL_ID in batch service list;Will The SQL_ID and the SQL_ID in batch service list that the session currently run performs sentence SQL are contrasted, and are known if matching Batch service is abnormal, is that unknown batch service is abnormal or the abnormal traffic is not batch service if mismatching;And
Information alert unit, for being sent a warning message according to recognition result to alarm platform.
9. device as claimed in claim 8, it is characterised in that the resource probe unit includes operating-system resources probe mould Block and/or Database Systems resource probe module;
The operating-system resources probe module, for detecting operating-system resources consumption;
The Database Systems resource probe module, for detection data storehouse system resources consumption;
The time reference line unit includes operating-system resources base line module and/or Database Systems resource base line module;
The operating-system resources base line module, for calculating operating-system resources baseline;
The Database Systems resource base line module, for calculating Database Systems resource baseline.
10. device as claimed in claim 9, it is characterised in that the Database Systems resource probe module, which includes statistics, to be believed Cease probe submodule and/or live traffic probe submodule;
The statistical information probe submodule, for detecting statistical information consumption;
The live traffic probe submodule, consumed for detected event business;
Statistical information consumption is average login times per second, average logic per second is read, average physical read per second, flat Number of transactions per second, average per second day quality, average active session per second, average the accumulative of rewind journal per second make Dosage, average hard parsing per second, and average vernier per second are opened at least one in number;
The live traffic consumption waits for roll-back segment contention, indexes contention wait, the wait of serial contention, focus block contention etc. Treat, and journal file is synchronously waiting at least one.
11. device as claimed in claim 9, it is characterised in that the operating-system resources base line module includes:
Activity duration splits submodule, for being segmented according to the batch service of the statistics execution time to the activity duration;
First mean value computation submodule, for calculating the operating-system resources consumption respectively in setting time threshold value daily The average M1 of average in each period, and in setting time threshold value within daily each period peak average M2, and calculate the average M3 of the average M1 and average M2.
12. device as claimed in claim 9, it is characterised in that the Database Systems resource base line module includes:
Second mean value computation submodule, for calculating the Database Systems resource consumption respectively in setting time threshold value every The average M4 of average in day each period, and in setting time threshold value within daily each period peak average M5, and calculate the average M6 of the average M4 and average M5.
13. device as claimed in claim 8, it is characterised in that information alert unit includes as follows:
Known batch service abnormality alarming module, for when the abnormal batch service is known batch service, described in judgement Whether abnormal batch service performs in regular time periods, is to send batch service property abnormality warning information to alerting platform, no Then send batch service abnormal time point and perform warning information to alerting platform;
Unknown batch service abnormality alarming module, for when the abnormal batch service is unknown batch service, sending unknown Service exception warning information is to alerting platform;
Newly-increased batch service alarm module, for when the new batch service of discovery, or the new spy of the existing batch service of discovery During sign, the new feature information of the addition new batch service or existing batch service is into batch service list;
Non- batch service abnormality alarming module is abnormal for when the abnormal traffic is unrelated with batch service, sending non-batch Warning information is put on record abnormal conditions addition document library to alerting platform.
CN201210423143.4A 2012-10-29 2012-10-29 A kind of batch service method for early warning and device Active CN103793309B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210423143.4A CN103793309B (en) 2012-10-29 2012-10-29 A kind of batch service method for early warning and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210423143.4A CN103793309B (en) 2012-10-29 2012-10-29 A kind of batch service method for early warning and device

Publications (2)

Publication Number Publication Date
CN103793309A CN103793309A (en) 2014-05-14
CN103793309B true CN103793309B (en) 2017-11-21

Family

ID=50669012

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210423143.4A Active CN103793309B (en) 2012-10-29 2012-10-29 A kind of batch service method for early warning and device

Country Status (1)

Country Link
CN (1) CN103793309B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104363113A (en) * 2014-10-29 2015-02-18 中国建设银行股份有限公司 Business continuity detection method
CN105589785A (en) * 2015-12-08 2016-05-18 中国银联股份有限公司 Device and method for monitoring IO (Input/Output) performance of storage equipment
CN105610647A (en) * 2015-12-30 2016-05-25 华为技术有限公司 Service abnormity detection method and server

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1677934A (en) * 2004-03-31 2005-10-05 华为技术有限公司 Method and system for monitoring network service performance
CN1794645A (en) * 2005-08-24 2006-06-28 上海浦东软件园信息技术有限公司 Invading detection method and system based on procedure action
CN101075919A (en) * 2006-06-22 2007-11-21 腾讯科技(深圳)有限公司 Method and system for monitoring Internet service
CN101123786A (en) * 2007-07-26 2008-02-13 中国移动通信集团山东有限公司 Intelligent control method for GRPS service
CN101692736A (en) * 2009-09-16 2010-04-07 南京联创科技集团股份有限公司 Method for monitoring telecom mobile service exchange based on flex technology

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101136758A (en) * 2007-07-20 2008-03-05 南京联创科技股份有限公司 Application method for online accounting system in owing risk control system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1677934A (en) * 2004-03-31 2005-10-05 华为技术有限公司 Method and system for monitoring network service performance
CN1794645A (en) * 2005-08-24 2006-06-28 上海浦东软件园信息技术有限公司 Invading detection method and system based on procedure action
CN101075919A (en) * 2006-06-22 2007-11-21 腾讯科技(深圳)有限公司 Method and system for monitoring Internet service
CN101123786A (en) * 2007-07-26 2008-02-13 中国移动通信集团山东有限公司 Intelligent control method for GRPS service
CN101692736A (en) * 2009-09-16 2010-04-07 南京联创科技集团股份有限公司 Method for monitoring telecom mobile service exchange based on flex technology

Also Published As

Publication number Publication date
CN103793309A (en) 2014-05-14

Similar Documents

Publication Publication Date Title
Laptev et al. Generic and scalable framework for automated time-series anomaly detection
US20200358826A1 (en) Methods and apparatus to assess compliance of a virtual computing environment
Chen et al. Non-parametric scan statistics for event detection and forecasting in heterogeneous social media graphs
CN105718351B (en) A kind of distributed monitoring management system towards Hadoop clusters
Jia et al. Prediction and analysis of Coronavirus Disease 2019
US9419917B2 (en) System and method of semantically modelling and monitoring applications and software architecture hosted by an IaaS provider
US9558347B2 (en) Detecting anomalous user behavior using generative models of user actions
US10038618B2 (en) System event analyzer and outlier visualization
Wong et al. Quantifying political leaning from tweets and retweets
US8850263B1 (en) Streaming and sampling in real-time log analysis
Zou et al. A docker container anomaly monitoring system based on optimized isolation forest
US9658916B2 (en) System analysis device, system analysis method and system analysis program
EP2871577B1 (en) Complex event processing (CEP) based system for handling performance issues of a CEP system and corresponding method
CN108052528B (en) A kind of storage equipment timing classification method for early warning
Salfner et al. A survey of online failure prediction methods
AU2018200016B2 (en) Systems and methods for anomaly detection
Long et al. A longitudinal survey of internet host reliability
CN104205063B (en) Operation administration device, operation administration method, and program
Schroeder et al. A large-scale study of failures in high-performance computing systems
Rolka et al. Issues in applied statistics for public health bioterrorism surveillance using multiple data streams: research needs
CN105357038B (en) Monitor the method and system of cluster virtual machine
CN103354924B (en) For monitoring the method and system of performance indications
US8732534B2 (en) Predictive incident management
US8086708B2 (en) Automated and adaptive threshold setting
US6629266B1 (en) Method and system for transparent symptom-based selective software rejuvenation

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
GR01 Patent grant