Difference between revisions of "PMQ"
From ESS-WIKI
| (36 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
| − | + | '''PMQ''' ('''Prediction Management Qualtity''') is a series of Mechine Learning and Big Data solutions for monitoring IPC/IOT Data in '''WISE-PaaS/RMM'''. Be diffierent with common device data monitor, it offers enhanced and multivariant function to help users reduce cost for monitoring tens of thousands IPC/IOT divices. PMQ not only aims at real-time or history data monitor, but also use these data to analysis effects for specific equipments ('''HDD Failure Analysis''') and predict device data in furture ('''Trend Forecasting Monitor''') by PMQ intelligent calculation in Devices. | |
| − | + | [[File:PMQ.jpg|480x344px|PMQ.jpg]] | |
| − | + | == <span style="font-size:large;">HDD F</span>ailure <span style="font-size:large;">Analysis</span> == | |
| − | < | + | ;<font size="3">Function</font> |
| + | :For device with HDD failure prediction intelligent application, users can know related status information of HDD that WISE-PaaS/RMM agent deliver. Status information contains predictive status of HDD, alert message and corresponding suggestion to help user check or maintain related equipments, then reduce situations that HDD fails in real. | ||
| + | ;<font size="3">Implementation</font> | ||
| + | :The major function of HDD failure prediction intelligent application in device is beased on some data of indicative factor (contain statistic of abnormal power off count, HDD write/read issue and so on) to give predictive status of HDD and alert messages. For predictive status of HDD, it uses build-in trained regression analysis model to predict for data of Indicative factor. For alert message, it defines a series of rules by practical test in advance, then these content of rules is given the reasonable range of indicative factors, and regards them as the standard of alert determination. According to two kind of informations, WISE-PaaS/RMM UI designs a scan button to enable this function , it will show analysis results in one hour for each device. When users focus on one in analysis results , users can know some suggestions for predictive status and alert meassage, and these suggestions are related with some equipments, so UI also shows these status of equipments (for instance, system CPU temperature and memory loading) for users to check. | ||
| − | + | == <span style="font-size:large;">Trend Forecasting Monitor</span> == | |
| − | < | + | ;<font size="3">Function</font> |
| − | Result for each object contains 3 parts: result status, status information and event record list. First, for result status. except | + | *Predict IPC/IOT Data in Future |
| + | *Automatically Monitor Alteration of IPC/IOT Data | ||
| + | |||
| + | :For sepecific device or sensor data (HDD Health, System CPU Usage and so on), support prediction application ('''Trend Forecasting Monitor''') in device. It can help users know abnormal data status as soon as possible to prevent some damages.Predictive Type of Trend Forecasting Monitor are '''hourly averages''' of data, and they contain averages of current hour and even next 3 hours in furtue.<br/> | ||
| + | :Otherwise, Trend Forecasting Monitor also offer function of automatical monitoring hourly averages of IPC/IOT data. If it happens situation that hourly averages or predictions change significantly, It will also notify users. | ||
| + | :<br/> | ||
| + | ;<span style="font-size:medium;">Predictive Process</span> | ||
| + | :The first condiction is predictive object has at least '''the newest 48 history hourly averages''' in Database, and these samples are calculated by time-series prediction algorithm. Time-series prediction algorithm '''Trend Forecasting Monitor '''adopts is '''Exponential Smoothing'''. By the above Process, it can produce values of prediction(averages of current hour and next 3 hours), and Trend Forecasting Monitor takes these values and the newest hourly average to compare with hourly averages in yesterday. If there are significant differences between values of prediction and hourly averages in yesterday, it will give alert message with potential risk. If there are significant differences between the newest hourly averages and hourly averages in yesterday, give alert message with doubtful risk. The standard of significant differences is based on '''Tukey's Range Test''', and the formula is: <span style="font-size:small;">'''[Q<sub>1</sub>-3(Q<sub>3</sub>-Q<sub>1</sub>), Q3+3(Q<sub>3</sub>-Q<sub>1</sub>)]'''. </span>Q<sub>1</sub> and Q<sub><span style="font-size: 12px;">3</span></sub> are the lower and upper quartile of hourly averages in yesterday. If data are in the above range, there are not significant differences between them, then the data status is normal. <br/> | ||
| + | :<br/> | ||
| + | ;<span style="font-size:medium;">Predictive Result</span> | ||
| + | :Result for each object contains 3 parts: result status, status information and event record list. First, for result status. except potential risk", doubtful risk and normal in the above statement, there is a situation abour data error, and it means disable predicting because of some data problem. For instance, miss some data in the newest 48 hours or query data from database error. Second, according to result status, web UI will show corresponded information, and major statement is explain reason for status and give some suggestions to users.Finally, event record list will show total records for significant change, then it can help users to trace initial time of abnormal data. | ||
Latest revision as of 09:42, 23 March 2017
PMQ (Prediction Management Qualtity) is a series of Mechine Learning and Big Data solutions for monitoring IPC/IOT Data in WISE-PaaS/RMM. Be diffierent with common device data monitor, it offers enhanced and multivariant function to help users reduce cost for monitoring tens of thousands IPC/IOT divices. PMQ not only aims at real-time or history data monitor, but also use these data to analysis effects for specific equipments (HDD Failure Analysis) and predict device data in furture (Trend Forecasting Monitor) by PMQ intelligent calculation in Devices.
HDD Failure Analysis
- Function
- For device with HDD failure prediction intelligent application, users can know related status information of HDD that WISE-PaaS/RMM agent deliver. Status information contains predictive status of HDD, alert message and corresponding suggestion to help user check or maintain related equipments, then reduce situations that HDD fails in real.
- Implementation
- The major function of HDD failure prediction intelligent application in device is beased on some data of indicative factor (contain statistic of abnormal power off count, HDD write/read issue and so on) to give predictive status of HDD and alert messages. For predictive status of HDD, it uses build-in trained regression analysis model to predict for data of Indicative factor. For alert message, it defines a series of rules by practical test in advance, then these content of rules is given the reasonable range of indicative factors, and regards them as the standard of alert determination. According to two kind of informations, WISE-PaaS/RMM UI designs a scan button to enable this function , it will show analysis results in one hour for each device. When users focus on one in analysis results , users can know some suggestions for predictive status and alert meassage, and these suggestions are related with some equipments, so UI also shows these status of equipments (for instance, system CPU temperature and memory loading) for users to check.
Trend Forecasting Monitor
- Function
- Predict IPC/IOT Data in Future
- Automatically Monitor Alteration of IPC/IOT Data
- For sepecific device or sensor data (HDD Health, System CPU Usage and so on), support prediction application (Trend Forecasting Monitor) in device. It can help users know abnormal data status as soon as possible to prevent some damages.Predictive Type of Trend Forecasting Monitor are hourly averages of data, and they contain averages of current hour and even next 3 hours in furtue.
- Otherwise, Trend Forecasting Monitor also offer function of automatical monitoring hourly averages of IPC/IOT data. If it happens situation that hourly averages or predictions change significantly, It will also notify users.
- Predictive Process
- The first condiction is predictive object has at least the newest 48 history hourly averages in Database, and these samples are calculated by time-series prediction algorithm. Time-series prediction algorithm Trend Forecasting Monitor adopts is Exponential Smoothing. By the above Process, it can produce values of prediction(averages of current hour and next 3 hours), and Trend Forecasting Monitor takes these values and the newest hourly average to compare with hourly averages in yesterday. If there are significant differences between values of prediction and hourly averages in yesterday, it will give alert message with potential risk. If there are significant differences between the newest hourly averages and hourly averages in yesterday, give alert message with doubtful risk. The standard of significant differences is based on Tukey's Range Test, and the formula is: [Q1-3(Q3-Q1), Q3+3(Q3-Q1)]. Q1 and Q3 are the lower and upper quartile of hourly averages in yesterday. If data are in the above range, there are not significant differences between them, then the data status is normal.
- Predictive Result
- Result for each object contains 3 parts: result status, status information and event record list. First, for result status. except potential risk", doubtful risk and normal in the above statement, there is a situation abour data error, and it means disable predicting because of some data problem. For instance, miss some data in the newest 48 hours or query data from database error. Second, according to result status, web UI will show corresponded information, and major statement is explain reason for status and give some suggestions to users.Finally, event record list will show total records for significant change, then it can help users to trace initial time of abnormal data.