|
Measures reported by EgThresholdManagerTest
The Threshold manager process computes the thresholds - i.e., the upper/limits of performance - for every measure collected by the eG agent and stores the thresholds so computed in the eG database. Since thresholds govern the state of a measure, if the threshold manager process fails to compute thresholds or computes them slowly or fails to even run, it can grossly impair the eG monitoring solution's ability to promptly detect problem areas. Using this test, you can understand how efficient the threshold manager process is. The test reports the current status of this process, points you to threshold computation failures and where they occurred, and reveals slowdowns in threshold computation (if any).
The measures made by this test are as follows:
| Measurement |
Description |
Measurement Unit |
Interpretation |
| Threshold_Status |
Indicates the current status of the threshold manager process. |
|
The values that this measure reports and the numeric values that correspond to them have been discussed in the table below:
| Measure Value |
Numeric Value |
| Running |
0 |
| Sleeping |
1 |
| Error |
2 |
| Not started |
3 |
| Interrupted |
4 |
Note:
By default, this measure reports the Measure Values listed in the table above to indicate the current status of the threshold manager process. The graph of this measure however, represents the same using the numeric equivalents only. |
| Time_Taken_For_Threshold |
Indicates the total time taken by the threshold manager to compute thresholds. |
Minutes |
Ideally, the value of this measure should be low. A steady rise in this measure value is a cause for concern, as it indicates that the threshold manager is taking too long to compute thresholds. This can happen if the threshold manager needs to compute thresholds for too many tests, measures, and descriptors. |
| Success_Threshold_Tests |
Indicates the number of tests for which thresholds were computed successfully. |
Number |
You can use the detailed diagnosis of this measure to know the tests for which thresholds have been successfully computed. |
| Failed_Threshold_Tests |
Indicates the number of tests for which threshold computation failed. |
Number |
The value 0 is desired for this measure. Any non-zero value is indicative of a thresholding failure. In this case, you can use the detailed diagnosis of this measure to identify those tests for which threshold computation failed and investigate the reason why. Without thresholds, the monitoring solution cannot detect problem conditions; nor can it compute state. |
| Time_Since_Last_Thresh |
Indicates the elapsed time since the last threshold computation. |
Minutes |
Typically, thresholding is scheduled to take place at the end of every day. By carefully observing the values reported by this measure, you can easily find out when a scheduled threshold computation cycle was missed. |
| Is_Thresh_Seperate_Proc |
Indicates whether/not the threshold manager is running as a separate process. |
|
The eG manager is a 32-bit application that runs as a Java process. The maximum heap memory that can be allocated to this process is limited to 1.5 GB. Even if the physical server on which the eG manager is installed has more memory, since it is a single Java process, the eG manager cannot exploit the additional memory available on the server. To overcome this limitation, in eG Enterprise, the critical eG manager functions such as email alert management, threshold computation, trending, and database cleanup activities can all be run as separate Java processes (i.e., in addition to the core eG manager process).
Removing these key functions from the core eG manager process makes additional memory available for the core eG manager functions including data reception and analysis, alarm correlation, and web-based access and reporting. This reconfiguration of the eG manager into separate Java processes allows the eG manager to make better utilization of available server hardware resources and thereby offers enhanced scalability. In turn, this allows customers to get more leverage from their existing investment in the hardware that hosts the eG manager.
If cleanup has been configured to run as a separate Java process, then the value of this measure will be Yes. If not, then this measure reports the value No.
The numeric values that correspond to the measure values above are as follows:
| Measure Value |
Numeric Value |
| Yes |
1 |
| No |
0 |
Note:
By default, this measure reports the Measure Values listed in the table above to indicate whether/not cleanup runs as a separate Java process. The graph of this measure however, represents the same using the numeric equivalents only. |
| Slow_Threshold_Tests |
Indicates the number of tests for which the threshold computation was slow. |
Number |
Use the detailed diagnosis of this emasure to identify the tests that were slow in computing the thresholds. |
|