CN103200039B - Data monitoring method and device - Google Patents
Data monitoring method and device Download PDFInfo
- Publication number
- CN103200039B CN103200039B CN201210004693.2A CN201210004693A CN103200039B CN 103200039 B CN103200039 B CN 103200039B CN 201210004693 A CN201210004693 A CN 201210004693A CN 103200039 B CN103200039 B CN 103200039B
- Authority
- CN
- China
- Prior art keywords
- data
- baseline
- monitored
- setting
- warning strategies
- 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
Links
Landscapes
- Alarm Systems (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a data monitoring method and device, wherein the method and the device are applied to electronic commerce. The data monitoring method includes the following steps that according to the types of data to be monitored of an electronic commerce system, one or more warning strategies are chosen, and the warning strategies comprise a fixed value warning strategy, a base line warning strategy and a fluctuation warning strategy, wherein the fixed value warning strategy is used for warning when the data to be monitored exceed a preset threshold value, the base line warning strategy is used for comparing the data to be monitored with data on a preset base line and warning when the data to be monitored exceed a preset range, and the fluctuation warning strategy is used for comparing the data to be monitored with preset data and warning when the data to be monitored exceed a preset range; and according to the chosen warning strategies, monitoring is carried out on the data to be monitored. According to the method and the device, an effect of flexible monitoring for the data of the electronic commerce system is achieved.
Description
Technical field
The application is related to networking technology area, more particularly to a kind of data monitoring method being applied in ecommerce and
Device.
Background technology
Ecommerce refers in the extensive trade activity in all parts of the world, under the open network environment in internet,
Based on browser/server application mode, both parties carry out various commercial activities with not meeting, and realize the online purchase of consumer
Online transaction between thing, trade company and online pay by mails and various commercial activity, transaction, finance activities and correlation
A kind of new commercial operation pattern of integrated service activity.At present, the application of ecommerce is worldwide just with surprising
Speed popularization with development.
Popularization with ecommerce and development, people are more next to the transaction data of the people participating in business activity or enterprise
More pay close attention to, true and reliable transaction data can help the both sides of business transaction to understand mutually, and contribute to reducing business transaction
Risk.For this reason, arising at the historic moment to the monitoring system that ecommerce is monitored.
Existing ecommerce monitoring system when being monitored, accused using threshold values i.e. fixed value by great majority
Alert, such as cpu, load etc. exceed how many just alarms, and in daily record, abnormal number exceedes how many just alarms etc..But, for friendship
(unique visitor accesses certain website or clicks on certain news pv (pageviews, page browsing amount), the uv of easy website
Different ip addresses number), transaction stroke count, the business-level data such as the amount of money, the tendency of its data institute line drawing is the work with people
Dynamic related.No matter it is fixed value alarm using the maximum of whole day or minimum of a value, be all unaccommodated.For example, it is
When system is out of order, data can be fallen, but not necessarily can fall and trough or crest value, and now alarm can be failed to report, the damage of one minute x ten thousand
Lose, enterprise holds and dares not accept, and customer loss is more fearful；For another example, night someone plays prestige or climbs data, can cause data wire
High, but be because that, less than crest value, problem will not be found, except unartificial 24 hours stare at screen, find data tendency not
Normally, then triggering alarm manually again.In addition, fixed value only adapts to the alarm configuration of certain time period, special time such as 1 year
Afterwards, business change, now, this fixed value no longer may have meaning, and needs to re-start modification etc..
It can be seen that, current ecommerce monitoring system monitor mode is single, for rule and/or accident consideration all
It is not enough it is impossible to there is accident, or effectively monitored when data rule changes.
Accordingly, it would be desirable to a technical problem of the urgent solution of those skilled in the art is exactly: how to comply with ecommerce and send out
Exhibition, is neatly monitored to the data of e-commerce system, even if in generation accident, or data rule changes
When, also ecommerce can be effectively monitored.
Content of the invention
The application provides a kind of data monitoring method and device, to solve the problems, such as the data monitoring of e-commerce system.
In order to solve the above problems, this application discloses a kind of data monitoring method, comprising: according to e-commerce system
The type of data to be monitored, selects one or more of warning strategies, described warning strategies include: fixed value warning strategies,
Baseline warning strategies and fluctuation warning strategies, wherein, described fixed value warning strategies are used for exceeding in described data to be monitored setting
Alerted during fixed threshold values, described baseline warning strategies are used for the data phase in described data to be monitored with the baseline of setting
Relatively, alerted during the scope exceeding setting, described fluctuation warning strategies are used in described data to be monitored and the number setting
According to comparing, alerted during the scope exceeding setting；According to the described warning strategies selecting, described data to be monitored is carried out
Monitoring.
Preferably, in the type of the described data to be monitored according to e-commerce system, one of warning strategies are selected
Or before multiple steps, also include: according to the previous data genaration baseline of the described e-commerce system in setting time；Root
According to the baseline of described generation, described baseline warning strategies are set.
Preferably, the step bag of the described previous data genaration baseline according to the described e-commerce system in setting time
Include: obtain the described previous data in setting time；According to setting rule, the described previous data obtaining is grouped, calculates
In the group of each group, mean value and group internal standard are poor；Using mean value and described group of internal standard in described group of each group described
Difference, filters to the data in this group according to setting rule, obtains retention data in the group of this group；Institute to each group described
State retention data in group to be averaging, obtain the initial baseline value of this group；All of described initial baseline value is smoothed, according to
Described initial baseline value after smooth generates described baseline.
Preferably, the step that the described initial baseline value after described basis smooths generates described baseline includes: after smoothing
Described initial baseline value drawn high according to the baseline growth rate setting, generate institute using the described initial baseline value after drawing high
State baseline.
Preferably, in the type of the described data to be monitored according to e-commerce system, one of warning strategies are selected
Or before multiple steps, also include: judge that described data to be monitored occurs the size of the possibility of data fluctuations amplitude hit,
According to the parameter of judged result setting described fluctuation warning strategies, described parameter includes described data to be monitored and exceedes described setting
Data scope, and described data to be monitored exceedes the number of times of the scope of data of described setting.
Preferably, the type of the described data to be monitored according to e-commerce system, select one of warning strategies or
Multiple steps includes: if described data to be monitored is business-level data, selects described baseline warning strategies and/or described
Fluctuation warning strategies, and, the data described to be monitored setting in described fluctuation warning strategies exceedes the data of described setting
The number of times of scope is single；Wherein, described business data packet purse rope station pv, website uv, transaction stroke count and dealing money；If described
Data to be monitored is system level data, then select described fixed value warning strategies and/or described fluctuation warning strategies, and,
Set the number that the data described to be monitored in described fixed value warning strategies and/or described fluctuation warning strategies exceedes described setting
According to the number of times of scope be multiple；Wherein, described system level data includes cpu data, internal storage data, load data and network
Data on flows.
In order to solve the above problems, disclosed herein as well is a kind of data monitoring device, comprising: selecting module, for root
According to the type of the data to be monitored of e-commerce system, select one or more of warning strategies, described warning strategies include:
Fixed value warning strategies, baseline warning strategies and fluctuation warning strategies, wherein, described fixed value warning strategies are used for described
Data to be monitored is alerted when exceeding the threshold values of setting, and described baseline warning strategies are used in described data to be monitored and setting
Baseline on data compare, alerted during the scope exceeding setting, described fluctuation warning strategies be used for wait to supervise described
Control data, compared with the data setting, is alerted during the scope exceeding setting；Monitoring module, for according to select
Warning strategies, are monitored to described data to be monitored.
Preferably, data monitoring device also includes: baseline generation module, in described selecting module according to described electronics
The type of the data to be monitored of business system, before selecting one or more of warning strategies, according to the institute in setting time
State the previous data genaration baseline of e-commerce system；First strategy setting module, for the baseline according to described generation, is arranged
Described baseline warning strategies.
Preferably, described baseline generation module includes: acquisition module, for obtaining the described past issue in setting time
According to；Computing module, for being grouped to the described previous data obtaining according to setting rule, calculates average in the group of each group
Value and group internal standard are poor；Filtering module, poor for mean value in described group using each group described and described group of internal standard, press
According to setting rule, the data in this group is filtered, obtain retention data in the group of this group；Averaging module, for described every
In described group of individual group, retention data is averaging, and obtains the initial baseline value of this group；Smooth generation module, for all of institute
State initial baseline value to be smoothed, described baseline is generated according to the described initial baseline value after smooth.
Preferably, described smooth generation module includes: Leveling Block, for putting down to all of described initial baseline value
Sliding；Draw high module, for will smooth after described initial baseline value drawn high according to the baseline growth rate setting, using drawing high
Described initial baseline value afterwards generates described baseline.
Compared with prior art, the application has the advantage that
The application selects different warning strategies by the data for different e-commerce systems it is achieved that according to electronics
The characteristic of the data of business system, is neatly monitored to the data of e-commerce system.For example, for pv, uv, transaction pen
The business-level data such as number, amount of money, cannot be carried out effective monitoring using fixed value alarm, then can be entered using other warning strategies
Row monitoring, such as baseline warning strategies or fluctuation warning strategies etc.；And, select warning strategies on the basis of, for such as cpu,
Offered load, network traffics etc. allow the data of abrupt transients, can also first configure the parameter in warning strategies, such as join parameter
It is set to after data to be monitored exceedes the certain number of times of alarm standard and alerts, then using the warning strategies after configuration.It can be seen that, lead to
Cross the application, effectively prevent existing ecommerce adopt fixed value alarm mode single it is impossible to occur accident, or number
When changing according to rule, data is carried out with the problem of effective monitoring, has reached and neatly the data of e-commerce system has been carried out
Monitoring, even if in generation accident, or when data rule changes, also can be supervised to e-commerce system effectively
The effect of control.
Brief description
Fig. 1 is a kind of flow chart of steps of the data monitoring method according to the embodiment of the present application one；
Fig. 2 is a kind of flow chart of steps of the data monitoring method according to the embodiment of the present application two；
Fig. 3 is a kind of flow chart of steps of the data monitoring method according to the embodiment of the present application three；
Fig. 4 is the schematic diagram that one of embodiment illustrated in fig. 3 carries out warning strategies setting；
Fig. 5 is that one of embodiment illustrated in fig. 3 carries out warning strategies setting for the change of festivals or holidays electronic commerce data
Schematic diagram；
Fig. 6 is a kind of structured flowchart of the data monitoring device according to the embodiment of the present application four.
Specific embodiment
Understandable for enabling the above-mentioned purpose of the application, feature and advantage to become apparent from, below in conjunction with the accompanying drawings and specifically real
Mode of applying is described in further detail to the application.
Embodiment one
With reference to Fig. 1, it illustrates a kind of flow chart of steps of the data monitoring method according to the embodiment of the present application one.
The data monitoring method of the present embodiment comprises the following steps:
Step s102: the type of the data to be monitored according to e-commerce system, select one of warning strategies or many
Individual.
Wherein, warning strategies include: fixed value warning strategies, baseline warning strategies and fluctuation warning strategies.In above-mentioned announcement
In whipping a horse on slightly, fixed value warning strategies are used for being alerted when data to be monitored exceedes the threshold values of setting；Baseline warning strategies
, alerted during the scope exceeding setting compared with the data on the baseline setting in data to be monitored；Fluctuation alarm
Strategy, is alerted during the scope exceeding setting in data to be monitored compared with the data setting.
The data type of e-commerce system can follow the conventional definitions of this area, comprising: business-level data (pv
Uv transaction stroke count dealing money), system level data (cpu load (load) memory (internal memory) traffic (network flow
Amount)), infrastructure service rank (index such as jboss- connection pool _ internal memory the index such as apache- handling capacity _ corresponding time
The indexs such as tomcat- handling capacity _ corresponding time the index such as database-connection number _ cache hit rate), application level data is (different
Constant interface interchange number of times the interface corresponding time) etc..Certainly, not limited to this, those skilled in the art can also be according to reality
Situation, carries out self-defined division to the type of the data of e-commerce system, such as according to the development of the data of e-commerce system
Rule is divided, or divides etc., the application couple according to sudden (or the claiming mutability) of the data of e-commerce system
This is not restricted.
When selecting warning strategies, monitoring data can be treated and select one or more warning strategies to be monitored, e.g.,
The same time period selects multiple warning strategies, or, select a kind of warning strategies a time period, select in another time period
In another kind of warning strategies.When selecting, those skilled in the art flexibly can select according to actual conditions.
Step s104: according to the warning strategies selecting, treat monitoring data and be monitored.
After choosing warning strategies, if data to be monitored exceedes alarm standard, alerted.That is, by alarm
Strategy is it is achieved that data monitoring to e-commerce system.
By the present embodiment, be different e-commerce systems data select different warning strategies it is achieved that according to
The characteristic of the data of e-commerce system, is neatly monitored to data, effectively prevent existing ecommerce using fixation
Value alarm mode single it is impossible to there is accident, or when data rule changes, the data of e-commerce system is entered
The problem of row effective monitoring, has reached and neatly the data of e-commerce system has been monitored, even if in generation accident,
Or when data rule changes, the effect that also ecommerce can be effectively monitored.
Embodiment two
With reference to Fig. 2, it illustrates a kind of flow chart of steps of the data monitoring method according to the embodiment of the present application two.
The data monitoring method of the present embodiment comprises the following steps:
Step s202: according to the target data of e-commerce system, fixed value warning strategies are set.
In this step, fixed value warning strategies can adopt existing set-up mode, including the maximum threshold values of setting and minimum
Threshold values etc..Fixed value warning strategies are commonly available to daily different time sections and exceed the monitor control index that threshold values just can alert, tendency
Relatively gently with monitor control index that user-association is little.Such as: the abnormal number in application level data, infrastructure service level
Database link number in other data database caches hit rate, the monitoring of the data such as disk space in system level data
And alarm.
Step s204: according to the previous data genaration baseline of the e-commerce system in setting time, baseline alarm is set
Strategy.
Baseline is the datum line being calculated according to specific policy, by itself and instant data institute line drawing put together into
Row compares, and can be used for alerting, trend prediction, the function such as abnormal conditions instant analysis.
At present, existing use baseline carries out alerting and the E-business applications of instant Problem judgment are little, and these are few
In some applications, its baseline accuracy rate is also of problems.Such as, the burr of baseline, the unsmooth, baseline of baseline are forbidden
Really, regular (festivals or holidays) and non-regularity event (network interrupts on a large scale, interim commemoration day etc.) leads to baseline chart to produce to join
Examination is inscribed.
It is that baseline alarm is applied to ecommerce, the present embodiment is previous according to the e-commerce system in setting time
Data genaration baseline.The general principle of its baseline algorithm is it is simply that to the sample data having certain rule in a large number, gone using standard deviation
Fall to deviate the too big burr data of each time point average, then average is used with forward, backward, circulation often adjacent xiIndividual point is yi
Secondary average to reach smooth effect.Now, data point institute line drawing trend is substantially identical with Future Data with tendency, but because
The height of n average line has been pulled low in addition it is also necessary to calculate a growth rate according to nearest Information And Historical situation.According to growth rate
The baseline of following special time period just can be drawn out with the data after smooth.The baseline being generated by this kind of mode, can paste
Reflect the possible situation of change of the data of future time section with cutting, thus being preferably monitored to ecommerce.
Conventional have nearest 3 monthly average smooth base, nearest 2 monthly average smooth base, nearest 1 monthly average smooth
Baseline.That is, according to the data genaration baseline of nearest 3 months, according to the data genaration baseline of nearest 2 months, according to nearest 1 month
Data genaration baseline etc..Certainly, it is also possible to configuration produces 1 year, the baseline of half a year if data is enough.
Except above-mentioned baseline generating mode, those skilled in the art in actual applications, can also be using other suitable
Baseline generating mode, e.g., seeks the mean value of the previous data in setting time, then filters out previous data according to this mean value
In data against regulation, further according to data genaration baseline after filtering etc., the application is not restricted to this.
After baseline generates, can set and exceed baseline certain limit (percentage or occurrence etc.) when data to be monitored
When, that is, alerted, formed baseline warning strategies.
Step s206: judge that data to be monitored occurs the size of the possibility of data fluctuations amplitude hit, according to judgement knot
Fruit setting fluctuation warning strategies.
Setting fluctuation warning strategies include arranging the parameter of fluctuation warning strategies, wherein, the parameter bag of fluctuation warning strategies
Include the scope that data to be monitored exceedes the data of setting, and data to be monitored exceedes the number of times of the scope of the data of setting.
Fluctuation refers to this sub-value compared with the value of set point number, the percentage deviation value obtaining.As, this sub-value and last time
The comparison of value, gained percentage deviation value；Or, the comparison of this sub-value and nearest 3 averages, gained percentage deviation value etc..
In the existing monitoring system using fluctuation alarm, do not account for the area of the absolute state of data moment state data
Not.That is, the fluctuating range of some data can occur acute variation suddenly, and other data then will not occur such change
Change.Such as, in ecommerce, for business-level data such as website pv, uv, transaction stroke count, dealing money, and cpu, load
It is different that (load), network traffics etc. allow the system level data of moment state.Young tiger such as cpu, load, network traffics
Height, it may be possible to certain application starts use in x thread such as x second within the of short duration time, now should not alert；And it is above-mentioned
Business datum, it is immediately to need to alert that young tiger obtains very high or very low.Accordingly, it would be desirable to for different data, setting is different
Fluctuation warning strategies.
In the present embodiment, first judge that data to be monitored occurs the size of the possibility of data fluctuations amplitude hit, i.e. first examine
Consider the absolute state of data moment state data, then fluctuation warning strategies are set further according to judged result.This setup is directed to
The characteristic of the data of above-mentioned e-commerce system, arranges different fluctuation warning strategies, i.e. for trade type website pv, uv, friendship
The alarm of the easy business-level data such as stroke count, dealing money, is compared with upper sub-value using this sub-value, once exceedes threshold values and just accuse
Alert；And cpu, load, network traffics etc. are allowed to the data of moment state, it is possible to use this sub-value and nearest 3 average ratio relatively,
Once exceed just alarm；Or this sub-value is compared with upper sub-value, continuous several times, detect just alarm of ging wrong for such as 2 times.
By the fluctuation warning strategies set-up mode of the present embodiment, can be for the spy of the data of different e-commerce systems
Property, using different warning strategies, it is to avoid non-appropriate warning, can more effectively ecommerce be monitored.
In fact, including fixed value alarm, fluctuation alarm, baseline alarm can be applied to all e-commerce system
The data of rank, and can freely set generation how many times alarm.Preferably, pv uv dealing money transaction stroke count, this
The index proposed arrangement of several business-levels becomes supplemented by baseline based on fluctuation and confirms that number of times is once, and the data of other ranks
Then according to using scene and environment, freely configuring warning strategies and number of times can be confirmed.
It should be noted that the execution of step s202, step s204 and step s206 can in no particular order sequentially.
Step s208: according to the type of data to be monitored, select fixed value warning strategies, baseline warning strategies and fluctuation
One or more of warning strategies warning strategies.
During configuration warning strategies, option and installment baseline warning strategies, fluctuation warning strategies, fixed value warning strategies on demand.
This 3 kinds of warning strategies on demand, can select one or more configurations simultaneously, and such as different time sections select with different alarm sides
Method；The same time period configures several alarm methods simultaneously, each method meet simultaneously or one meet just alarm (even condition and with
Conditional relationship).
For example, for the monitoring of business datum pv, uv, transaction data etc., using traditional fixed value alarm it is impossible to solve
Produced problem between trough and crest, now can be using fluctuation alarm or baseline alarm.But, for the smooth decline of data
Or rising problem, fluctuation alarm is to solve, and baseline alarm now may be selected.By selecting different alarm modes, permissible
Solution is failed to report, understatement the problems such as.
By research find, actually used in, fluctuation alarm can solve most of alarm.Especially there iing 7*24 hour
On duty stare at screen in the case of, configure warning strategies when, for business datum can using fluctuation alarm based on, baseline alerts and is
Auxiliary scheme.
Step s210: according to the warning strategies selecting, treat monitoring data and be monitored.
By the present embodiment, the characteristic to dissimilar e-commerce system data, and, e-commerce system data
The possibility of rule and accident is fully considered, and then is provided with different warning strategies so that carrying out electricity
During son commercial affairs monitoring, it is possible to use fixed value (threshold values) alarm, baseline alarm, three kinds of alarm schemes of fluctuation alarm, according to difference
Scene suitably selects different warning strategies, is flexibly effectively realized ecommerce monitoring.For example, for business datum change or
Person malicious user climbs the scene of data night, can be alerted using the baseline more more particularly suitable than fixed value alarm or fluctuation,
Different time sections mix different threshold values, such as at night because data is less it is allowed to deviate baseline a little greatly or allow fluctuation width
More greatly, then triggering alerts degree.By the present embodiment, it is that monitoring warning system provides more accurately warning strategies, prevent by mistake
Report, fails to report；Provide more accurate trend prediction, provide foundation for decision-making in advance；Assist immediately to judge the abnormal feelings of data
Condition.
Embodiment three
With reference to Fig. 3, it illustrates a kind of flow chart of steps of the data monitoring method according to the embodiment of the present application three.
The data monitoring method of the present embodiment comprises the following steps:
Step s302: setting fixed value warning strategies.
That is, set a threshold values, the data to be monitored of e-commerce system is compared with the threshold values of this setting, is exceeding this valve
Alerted during value.The type of threshold values can be numeric type can also be character string, comparison condition include "=", "！=",
" ＞ ", " ＞=", " ＜ ", " ＜=", " containing ", " not containing " etc..
Step s304: setting fluctuation warning strategies.
That is, obtain the previous data of setting time point or the e-commerce system of time period, make data to be monitored and acquisition
Previous data be compared, exceed and alerted during certain limit.
In the present embodiment, the fluctuation warning strategies of setting include: data to be monitored is compared with previous point and front 3 points
Average specific compared with the point of time with last week compared with comparing with the point of time with yesterday.The configurable model that up or down fluctuates
Enclose, fluctuation range can be configured to percent amplitude or occurrence.
Step s306: setting baseline warning strategies.
Generate baseline warning strategies in baseline when, can by timed task read from repository baseline generation join
Put, according to the previous data genaration baseline in setting time.In the present embodiment, the baseline in baseline warning strategies includes: 3 months
Average smooth baseline, 2 monthly average smooth base, 1 monthly average smooth base.Configurable is higher or lower than baseline how many scopes
Shi Jinhang alerts, and the value of configuration can be percentage or occurrence.
Taking generate following 24 little base lines based on nearest 3 monthly average smooth base as a example, its product process includes:
Step s3062: in 24 hours taking out from nearest three months initial data and calculating the moment, there is identical work
Make all data of day+hour+minute combination, and it is pressed with working day+hour+minute packet, in group, be averaging u and standard
Difference ＆, obtains middle interim table tmptable.
It should be noted that for the data seeking baseline, the storage in object library needs time field.The present embodiment
In data be to preserve that is to say, that the number of minutes of every data can be all 3 multiple according to 3 minutes points, deposit within one hour
20 points, deposit 480 points for 24 hours.In addition, baseline generates time suggestion is placed on the data trough stage.
Understand under this situation, through being averaging u and standard deviation ＆ in group, total in the interim table tmptable of the centre obtaining
Article 480, record.
Certainly, in actual applications, those skilled in the art can be according to actual conditions, using other suitable packet sides
Formula is grouped to previous data, then calculates mean value and group internal standard in the group of each group poorer.
Step s3064: take out nearest 3 months data, filtered out according to the standard deviation in middle interim table and mean value u
There is the abnormal data in identical working day+hour+minute condition data, value x of the data of reservationiIt must is fulfilled for (u-＆) ＜ xi
＜ (u+1.5*＆).For the value remaining according still further to working day+hour+minute packet, obtain mean value in group and be initially
Baseline value, is now 480 points in the present embodiment.
In actual applications, those skilled in the art can be according to actual conditions, using other suitable rule-based filtering numbers
According to obtaining the data remaining, and then carry out packet and be averaging obtaining initial baseline value, the application is not restricted to this.
Step s3066: to initial baseline value, from front to back and each from back to front smooth 3 times: positive for the first time, reverse institute
There is adjacent 2 original value to be averaging and be assigned to first value；Second positive, reversely all 3 adjacent original values be averaging
It is assigned to first value；Positive for the third time, reversely all 4 adjacent original values be averaging and be assigned to first value.
By this step, the baseline that tendency and trend substantially and very smooth can be obtained, but the height of baseline is drawn
Low.
It should be noted that in this step, smooth mode is carried out to initial baseline value being merely illustrative, this area
Technical staff is in actual applications, it would however also be possible to employ other arbitrarily suitable smooth manners smooth to initial baseline value, this
Application is not restricted to this.
Step s3068: calculate today of this moment and last week in totally two days initial data (if 0 point of race base
Line production routine, then actually only have data today last week), take out today in calculate time-division work at present day in moment and it
Front all data to zero point, further take out all data after calculating today last week from the moment, being connected in series is exactly complete 24
Hour totally 480 data, and ask data and latestsum, take out work at present day totally 480 data from smooth rear data,
And ask data and smoothsum.Using the value of latestsum/smoothsum as following 24 little base line growth rates.
Certainly, not limited to this, baseline growth rate can also be set by the way of relatively simple, such as those skilled in the art
Directly set according to practical experience etc., the application is not restricted to this.
Step s30610: using baseline growth rate and smooth after 480 point-renderings go out the baseline of following 24 hours.
So far, completed based on nearest 3 monthly average smooth base of nearest 3 months data.
It is then possible to the baseline domain of walker being allowed according to this baseline, setting, generate and smoothed based on nearest 3 monthly average
The baseline warning strategies of baseline.
It should be noted that the execution of step s302, step s304 and step s306 can in no particular order sequentially.In addition,
Above-mentioned warning strategies, in setting, can be set according to the good rule of configured in advance, such as in certain time period to certain species
The data of type, using configured certain warning strategies of parameter, or it is also possible to by artificial according to actual conditions needs, hand
Dynamic flexibly setting.
Step s308: according to the type of data to be monitored, select warning strategies.
When selecting warning strategies, select baseline alarm, fluctuation alarm, fixed value alarm on demand.This 3 kinds alarm modes are on demand
Configuration, can select one or more configurations simultaneously, and such as different time sections select with different alarm methods；The same time period is same
When configure several alarm methods, each method meet simultaneously or one meet and just alert (even condition and and conditional relationship).Configuration
When can select on demand once or continuously to meet threshold values condition for x time, just inspire alarm.
A kind of for pv data carry out warning strategies setting schematic diagram as shown in Figure 4.Wherein, " threshold values type " is used for selecting
Select the particular type in warning strategies, and this warning strategies；" currency " is used for arranging the change of acceptable data to be monitored
Dynamic scope；" effective period of time " is used for arranging the time period to be monitored；" wrap count " is used for arranging acceptable number to be monitored
According to the number of times exceeding alarm threshold value.Figure 4, it is seen that being directed to pv data, it is provided with baseline warning strategies and fluctuation simultaneously
Warning strategies.
In addition, it is necessary to arrange solution for festivals or holidays or accident.Burst thing in festivals or holidays or target zone
Part, actual operating data can be led to deviate, and baseline ratio is more serious, and actual motion numerical value super large is extra small simultaneously, also results in fluctuation and accuses
The wrong report of alert and fixed value alarm or fail to report.
For this reason, for the festivals or holidays of fixing rule, baseline can be preset and declines adjustment it is also possible to not decline baseline, but
Original threshold values is carried out with a percentage overlap-add procedure.The deviation baseline 30% of such as original configuration alerts it now is possible to fold
Plus reach 50% just alarm.And fluctuating range can also be added to for 50% alarm for fluctuation alarm.And for numeric type
Fixed value is it is also possible to reduce percentage on demand.
And for accident, such as do promotion on Internet, mourn for activity and lead to that number of netizens increases sharply, network paralyses on a large scale
Deng it is also possible to according to above-mentioned festivals or holidays mode, on a time period meet an urgent need provisional configuration.One kind is directed to festivals or holidays e-commerce system
The schematic diagram that data variation carries out warning strategies setting is as shown in Figure 5.From figure 5 it can be seen that this alarm configuration is directed to
Pv data in business datum, is provided with baseline warning strategies for it, and, by being superimposed ratio, baseline warning strategies is carried out
Adjustment, so that it meets festivals or holidays ecommerce monitoring needs.
Step s310: according to the warning strategies selecting, treat monitoring data and be monitored.
By the present embodiment, according to the feature of different e-commerce system data, it is provided with different warning strategies, realize
Flexibly ecommerce monitoring effectively.The data monitoring method of the present embodiment effectively prevent existing ecommerce using fixation
Value alarm mode single it is impossible to there is accident, or when data rule changes, effective monitoring is carried out to ecommerce
Problem, reached and neatly ecommerce be monitored, even if there is accident, or data rule changed
When, the effect that also ecommerce can be effectively monitored.
Example IV
With reference to Fig. 6, it illustrates a kind of structured flowchart of the data monitoring device according to the embodiment of the present application four.
The data monitoring device of the present embodiment includes: selecting module 402, for the number to be monitored according to e-commerce system
According to type, select one or more of warning strategies, warning strategies include: fixed value warning strategies, baseline warning strategies,
With fluctuation warning strategies, wherein, fixed value warning strategies are used for being alerted when data to be monitored exceedes the threshold values of setting, base
Line warning strategies are used in data to be monitored, compared with the data on the baseline setting, being accused during the scope exceeding setting
Alert, fluctuation warning strategies are used in data to be monitored, compared with the data setting, being alerted during the scope exceeding setting；Prison
Control module 404, for according to the warning strategies selecting, treating monitoring data and being monitored.
Preferably, the data monitoring device of the present embodiment also includes: baseline generation module 406, in selecting module 402
The type of the data to be monitored according to e-commerce system, before selecting one or more of warning strategies, during according to setting
The previous data genaration baseline of interior e-commerce system；First strategy setting module 408, for the baseline according to generation,
Setting baseline warning strategies.
Preferably, baseline generation module 406 includes: acquisition module 4062, for obtaining the previous data in setting time；
Computing module 4064, for being grouped to the previous data obtaining according to setting rule, calculates mean value in the group of each group
Poor with group internal standard；Filtering module 4066, poor for mean value in the group organized using each and group internal standard, regular according to setting
Data in this group is filtered, obtains retention data in the group of this group；Averaging module 4068, in group that each is organized
Retention data is averaging, and obtains the initial baseline value of this group；Smooth generation module 40610, for all of initial baseline value
Smoothed, described baseline is generated according to the initial baseline value after smooth.
Preferably, smooth generation module 40610 includes: Leveling Block, for smoothing to all of initial baseline value；
Draw high module, drawn high according to the baseline growth rate setting for the initial baseline value after smoothing, using first after drawing high
Beginning baseline value generates baseline.
Preferably, the data monitoring device of the present embodiment also includes: judge module 410, in selecting module 402 basis
The type of the data to be monitored of e-commerce system, before selecting one or more of warning strategies, judges data to be monitored
There is the size of the possibility of data fluctuations amplitude hit；Second strategy setting module 412, for according to judge module 410
The parameter of judged result setting fluctuation warning strategies, described parameter includes the scope that data to be monitored exceedes the data of setting, and
Data to be monitored exceedes the number of times of the scope of the data of setting.
Preferably, if data to be monitored is business-level data, selecting module 402 selects baseline warning strategies for it
And/or fluctuation warning strategies, and, the data to be monitored setting in fluctuation warning strategies exceedes the secondary of the scope of data of setting
Number is single；If data to be monitored is system level data, select fixed value warning strategies and/or fluctuation warning strategies, and
And, the data to be monitored in setting fixed value warning strategies and/or fluctuation warning strategies exceedes the secondary of the scope of data of setting
Number is multiple；Wherein, business data packet purse rope station pv, website uv, transaction stroke count and dealing money；System level data includes
Cpu data, internal storage data, load data and network flow data.
The data monitoring device of the present embodiment is used for realizing corresponding data monitoring method in aforesaid plurality of embodiment of the method,
And there is the beneficial effect of corresponding embodiment of the method, will not be described here.
It is provided with multiple warning strategies, for the data of different e-commerce systems in the data monitoring scheme of the application
Type neatly can select different warning strategies, thus realizing the effective monitoring of ecommerce.And, to ecommerce system
The data of system has carried out the differentiation too of moment state and non-moment such that it is able to preferably configuring and selecting corresponding warning strategies.
By the data monitoring scheme of the application, the degree of accuracy and the instantaneity of monitoring alarm can be improved, help business forcast or instant
Pinpoint the problems.
Each embodiment in this specification is all described by the way of going forward one by one, what each embodiment stressed be with
The difference of other embodiment, between each embodiment identical similar partly mutually referring to.For device embodiment
For, due to itself and embodiment of the method basic simlarity, so description is fairly simple, referring to the portion of embodiment of the method in place of correlation
Defend oneself bright.
Above a kind of data monitoring method provided herein and device are described in detail, used herein
Specific case is set forth to the principle of the application and embodiment, and the explanation of above example is only intended to help understand this
The method of application and its core concept；Simultaneously for one of ordinary skill in the art, according to the thought of the application, concrete
All will change on embodiment and range of application, in sum, this specification content should not be construed as to the application's
Limit.
Claims (9)
1. a kind of data monitoring method is it is characterised in that include:
Judge that data to be monitored occurs the size of the possibility of data fluctuations amplitude hit, according to judged result setting fluctuation alarm
The parameter of strategy, described parameter includes the scope that described data to be monitored exceedes the data of setting, and described data to be monitored surpasses
Cross the number of times of the scope of the data of described setting；Described data to be monitored includes business-level data and system level data；Institute
State judged result and include the absolute state of data moment state data；
The type of the data to be monitored according to e-commerce system, selects one or more of warning strategies, described alarm plan
Slightly include: fixed value warning strategies, baseline warning strategies and fluctuation warning strategies, wherein, described fixed value warning strategies are used for
Alerted when described data to be monitored exceedes the threshold values of setting, described baseline warning strategies are used in described data to be monitored
Compared with the data on the baseline setting, alerted during the scope exceeding setting, described fluctuation warning strategies are used in institute
State data to be monitored compared with the data of described setting, alerted during the scope exceeding described setting；
According to the described warning strategies selecting, described data to be monitored is monitored.
2. method according to claim 1 is it is characterised in that in the described data to be monitored according to e-commerce system
Type, before selecting the step of one or more of warning strategies, also includes:
Previous data genaration baseline according to the described e-commerce system in setting time；
According to the baseline of described generation, described baseline warning strategies are set.
3. method according to claim 2 it is characterised in that described according to the described e-commerce system in setting time
The step of previous data genaration baseline include:
Obtain the described previous data in setting time；
According to setting rule, the described previous data obtaining is grouped, calculates mean value and group internal standard in the group of each group
Difference；
Poor using mean value in described group of each group described and described group of internal standard, according to setting rule to the data in this group
Filtered, obtained retention data in the group of this group；
Retention data in described group of each group described is averaging, obtains the initial baseline value of this group；
All of described initial baseline value is smoothed, described baseline is generated according to the described initial baseline value after smooth.
4. method according to claim 3 is it is characterised in that the described initial baseline value after described basis smooths generates institute
The step stating baseline includes:
Described initial baseline value after will be smooth is drawn high according to the baseline growth rate setting, using described initial after drawing high
Baseline value generates described baseline.
5. method according to claim 1 is it is characterised in that the class of the described data to be monitored according to e-commerce system
Type, selects the step of one or more of warning strategies to include:
If described data to be monitored is business-level data, select described baseline warning strategies and/or described fluctuation alarm plan
Omit, and, the data described to be monitored setting in described fluctuation warning strategies exceedes the number of times of the scope of data of described setting
For single；Wherein, described business data packet purse rope station pv, website uv, transaction stroke count and dealing money；
If described data to be monitored is system level data, select described fixed value warning strategies and/or described fluctuation alarm
Strategy, and, the data described to be monitored setting in described fixed value warning strategies and/or described fluctuation warning strategies exceedes institute
The number of times of scope stating the data setting is as repeatedly；Wherein, described system level data includes cpu data, internal storage data, load
Data and network flow data.
6. a kind of data monitoring device is it is characterised in that include:
Judge module, for judging the size of the possibility of data generation data fluctuations amplitude hit to be monitored；
Second strategy setting module, for the parameter of the judged result setting fluctuation warning strategies according to judge module, described ginseng
Number includes the scope that data to be monitored exceedes the data of setting, and data to be monitored exceedes the number of times of the scope of the data of setting；
Described data to be monitored includes business-level data and system level data；Described judged result includes data moment state data
Definitely state；
Selecting module, for the type of the data to be monitored according to e-commerce system, selects one of warning strategies or many
Individual, described warning strategies include: fixed value warning strategies, baseline warning strategies and fluctuation warning strategies, wherein, described fixation
Value warning strategies are used for being alerted when described data to be monitored exceedes the threshold values of setting, and described baseline warning strategies are used for
Described data to be monitored, compared with the data on the baseline setting, is alerted during the scope exceeding setting, and described fluctuation is accused
Whip a horse on and be slightly used in described data to be monitored, compared with the data of described setting, being accused during the scope exceeding described setting
Alert；
Monitoring module, for according to the described warning strategies selecting, being monitored to described data to be monitored.
7. device according to claim 6 is it is characterised in that also include:
Baseline generation module, selects according to the type of the data to be monitored of described e-commerce system in described selecting module
Before selecting one or more of warning strategies, according to the previous data genaration base of the described e-commerce system in setting time
Line；
First strategy setting module, for the baseline according to described generation, arranges described baseline warning strategies.
8. device according to claim 7 is it is characterised in that described baseline generation module includes:
Acquisition module, for obtaining the described previous data in setting time；
Computing module, for being grouped to the described previous data obtaining according to setting rule, calculates flat in the group of each group
Average and group internal standard are poor；
Filtering module, poor for mean value in described group using each group described and described group of internal standard, regular according to setting
Data in this group is filtered, obtains retention data in the group of this group；
Averaging module, is averaging for retention data in described group to each group described, obtains the initial baseline value of this group；
Smooth generation module, for smoothing to all of described initial baseline value, according to the described initial baseline after smoothing
Value generates described baseline.
9. device according to claim 8 is it is characterised in that described smooth generation module includes:
Leveling Block, for smoothing to all of described initial baseline value；
Draw high module, for will smooth after described initial baseline value drawn high according to the baseline growth rate setting, using drawing
Described initial baseline value after height generates described baseline.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210004693.2A CN103200039B (en) | 2012-01-09 | 2012-01-09 | Data monitoring method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210004693.2A CN103200039B (en) | 2012-01-09 | 2012-01-09 | Data monitoring method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103200039A CN103200039A (en) | 2013-07-10 |
| CN103200039B true CN103200039B (en) | 2017-01-18 |
Family
ID=48722405
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210004693.2A Active CN103200039B (en) | 2012-01-09 | 2012-01-09 | Data monitoring method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103200039B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106970846A (en) * | 2017-03-15 | 2017-07-21 | 北京驰波信息工程有限公司 | Payment system message is controlled and processing method, device |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572391B (en) * | 2013-10-16 | 2019-03-15 | 深圳市腾讯计算机系统有限公司 | Monitoring alarm tactics configuring method and device, monitoring alarm method and device |
| CN104852810B (en) * | 2014-02-18 | 2018-11-30 | 中国移动通信集团公司 | A kind of determination method and apparatus of business platform exception |
| CN104462445A (en) * | 2014-12-15 | 2015-03-25 | 北京国双科技有限公司 | Webpage access data processing method and webpage access data processing device |
| CN107306200B (en) * | 2016-04-22 | 2020-08-25 | 中国电信股份有限公司 | Network fault early warning method and gateway for network fault early warning |
| CN107545187A (en) * | 2016-06-24 | 2018-01-05 | 阿里巴巴集团控股有限公司 | A kind of processing method and processing device of service data |
| CN107705149A (en) * | 2017-09-22 | 2018-02-16 | 平安科技（深圳）有限公司 | Data method for real-time monitoring, device, terminal device and storage medium |
| CN108520318A (en) * | 2018-04-02 | 2018-09-11 | 深圳前海桔子信息技术有限公司 | A kind of method for safety monitoring, device, server and storage medium |
| CN108965055B (en) * | 2018-07-17 | 2021-07-13 | 成都信息工程大学 | Network flow abnormity detection method based on historical time point taking method |
| CN110764975B (en) * | 2018-07-27 | 2021-10-22 | 华为技术有限公司 | Early warning method and device for equipment performance and monitoring equipment |
| CN109254895A (en) * | 2018-08-21 | 2019-01-22 | 山东超越数控电子股份有限公司 | A kind of high-performance server accident analysis prediction technique based on BMC |
| CN110348718B (en) * | 2019-06-28 | 2023-11-14 | 北京淇瑀信息科技有限公司 | Service index monitoring method and device and electronic equipment |
| CN110362455B (en) * | 2019-07-15 | 2022-12-20 | 北京奇艺世纪科技有限公司 | Data processing method and data processing device |
| CN113495909A (en) * | 2020-04-06 | 2021-10-12 | 中信百信银行股份有限公司 | Customer complaint single quantity early warning method and device, electronic equipment and storage medium |
| CN113064569A (en) * | 2021-04-01 | 2021-07-02 | 银清科技有限公司 | Page visualization system, method and device |
| TWI789219B (en) * | 2022-01-21 | 2023-01-01 | 友訊科技股份有限公司 | Monitoring control assisting and leading method for network devices, its terminal equipment and readable storage medium |
| CN114844762B (en) * | 2022-04-19 | 2023-11-07 | 中南大学 | Alarm authenticity detection method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6633834B2 (en) * | 2001-08-29 | 2003-10-14 | Hewlett-Packard Development Company, L.P. | Baselining of data collector data |
| US6658367B2 (en) * | 2001-03-28 | 2003-12-02 | Hewlett-Packard Development Company, L.P. | System for time-bucketing of baselined data collector data |
| US7281041B2 (en) * | 2002-10-31 | 2007-10-09 | Hewlett-Packard Development Company, L.P. | Method and apparatus for providing a baselining and auto-thresholding framework |
| CN101267362A (en) * | 2008-05-16 | 2008-09-17 | 亿阳信通股份有限公司 | A dynamic identification method and its device for normal fluctuation range of performance normal value |
-
2012
- 2012-01-09 CN CN201210004693.2A patent/CN103200039B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6658367B2 (en) * | 2001-03-28 | 2003-12-02 | Hewlett-Packard Development Company, L.P. | System for time-bucketing of baselined data collector data |
| US6633834B2 (en) * | 2001-08-29 | 2003-10-14 | Hewlett-Packard Development Company, L.P. | Baselining of data collector data |
| US7281041B2 (en) * | 2002-10-31 | 2007-10-09 | Hewlett-Packard Development Company, L.P. | Method and apparatus for providing a baselining and auto-thresholding framework |
| CN101267362A (en) * | 2008-05-16 | 2008-09-17 | 亿阳信通股份有限公司 | A dynamic identification method and its device for normal fluctuation range of performance normal value |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106970846A (en) * | 2017-03-15 | 2017-07-21 | 北京驰波信息工程有限公司 | Payment system message is controlled and processing method, device |
| CN106970846B (en) * | 2017-03-15 | 2019-08-16 | 银清科技(北京)有限公司 | The control of payment system message and processing method, device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103200039A (en) | 2013-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103200039B (en) | Data monitoring method and device | |
| CN106100937A (en) | System monitoring method and apparatus | |
| Kremer et al. | Do random errors explain newsvendor behavior? | |
| CN107153593A (en) | A kind of Internet service monitoring threshold value determination method and device | |
| Shmueli et al. | Visualizing online auctions | |
| CN109726833A (en) | Dynamic adjustment maintenance policy method, apparatus, terminal and computer storage medium | |
| CN106844673A (en) | A kind of method and system based on the public security data acquisition intimate degree of multidimensional personnel | |
| CN110138688A (en) | Dynamic adjusts method, apparatus, equipment and the readable storage medium storing program for executing of business interface | |
| CN107094207A (en) | A kind of traffic indicator abnormal automatic testing method and its automatic detection device | |
| Hillemeier et al. | Relative or absolute standards for child poverty: a state-level analysis of infant and child mortality | |
| Wang | Coordinating independent buyers with integer‐ratio time coordination and quantity discounts | |
| CN114186855A (en) | Monitoring and early warning method, device, computer equipment, storage medium and program product | |
| Guegan et al. | A mathematical resurgence of risk management: an extreme modeling of expert opinions | |
| WO2014150368A2 (en) | Systems, devices, and methods for generating location establishment severity identification | |
| US11983774B1 (en) | Systems and methods for determining a fixed rate based on behavior | |
| Quinn et al. | Envisioning the Smart Grid: Network architecture, information control, and the public policy balancing act | |
| Rossi et al. | Early warning of drought: development of a drought bullettin for Sicily | |
| CN106022920A (en) | Financial market pricing and quoting platform | |
| CN107301484A (en) | A kind of control point method to set up and device | |
| CN110059906B (en) | Policy effectiveness analysis method, device, server and storage medium | |
| Conceicão et al. | Brief on discounting in the context of climate change economics | |
| CN106408331A (en) | Message sending method and device | |
| CN107193845A (en) | Internet resources screening technique and device, the qualified assets screening technique of financing plans | |
| Ward et al. | Economic impacts of instream flow protection for the Rio Grande silvery minnow in the Rio Grande basin | |
| Cheung et al. | Balanced and synchronized ordering in supply chains |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1182860 Country of ref document: HK |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: GR Ref document number: 1182860 Country of ref document: HK |