Business Context
An IT Operations center monitors server error logs every hour. Due to the high volume of data (24/7 monitoring), analysts do not want to see 'normal' charts. They strictly want a report generated ONLY when the error count exceeds the control limits (Exceptions), and they want the results saved to a table for a dashboard feed.
Data Preparation
Simulating 48 hours of server error counts. Most hours are stable (low errors), but hour 35 has a massive spike due to a simulated outage.
Copied!
| 1 | |
| 2 | DATA mycas.ServerLogs; |
| 3 | DO i=1 to 48; |
| 4 | HourID=i; |
| 5 | IF i=35 THEN ErrorCount=150; |
| 6 | ELSE ErrorCount=10 + floor(rand('uniform')*5); |
| 7 | OUTPUT; |
| 8 | END; |
| 9 | |
| 10 | RUN; |
| 11 |
Étapes de réalisation
1
Execute cChart with 'exChart' set to True to suppress normal output, and save summary to 'DashboardData'.
Copied!
| 1 | |
| 2 | PROC CAS; |
| 3 | spc.cChart / TABLE={name='ServerLogs'} processValue='ErrorCount' subgroupValue='HourID' exChart=true limitN=1 chartsTable={name='DashboardData', replace=true}; |
| 4 | |
| 5 | RUN; |
| 6 |
Expected Result
The action should NOT display a visual chart for the stable periods. It should produce output/visualization specifically focusing on the exception at Hour 35. The 'mycas.DashboardData' table should be created containing the control limits and the flagged exception data.