Implementing Oracle Self-monitoring with metrics baseline technology

For example, the database administrator now needs to have the Oracle database System Monitor itself. Automatically alert the database administrator when the CPU usage of the server is reached, or automatically remind the database administrator, etc. when the usage of the tablespace reaches 90% of the full maximum capacity. These features are certainly useful for database administrators. There are several ways to help administrators monitor Oracle application environments in Enterprise Manager for Oracle database systems. The Oracle Management agent on each monitoring host can monitor the status, performance, and health of all targets on the host. If the target object fails or the performance reaches the early warning threshold, the database system generates early warning and tells the database administrator through certain channels. This database system can be seen as a self-monitoring function, which could replace the database administrator's eyes.

However, this self-monitoring function is dependent on the metric baseline. A metric baseline is a specification that tells the Enterprise Manager to perform an action when an indicator reaches the baseline of the measure or exceeds the number. Therefore, the management of metrics baselines is critical to the self-monitoring capabilities of the database system. If this metric baseline is set high, then the monitoring will not work. The database may have crashed when the system was alerted. If this metric baseline is set relatively low, then the database is frequently reported as invalid. Instead, it is too sensitive to cause unnecessary trouble to the database administrator. So the measurement baseline setting is critical for Oracle databases.

First, the basic steps to measure baseline settings

Setting a measurement baseline is not very difficult. Only a few simple steps are required to complete the measurement baseline settings. The difficulty is how to set the reasonable. The following article discusses how to set a reasonable measurement baseline based on the content of the measurement baseline settings.

Enter Oracle's Enterprise Manager to open the page sign that creates the measurement baseline. On this page check, you can set up all the contents of the measurement baseline. The database system will then automatically calculate the thresholds based on the settings here. When you create a measurement baseline, you select the date and time period in which the target performance is under a typical workload under the gram acceptance, which is the basic guideline for establishing a reasonable measurement baseline. Specifically, to set a measurement baseline, you need to set the specific parameters.

The first is the name of the metric baseline. Because the database can monitor multiple database objects and Oracle application environments at the same time, you need to give the metric baseline a proper name. This is important for subsequent maintenance of the database administrator. Do well in the name of the baseline to reflect the name of the monitored object. So the database administrator can know at a glance what this metric baseline is primarily used to monitor what objects. In addition, the database administrator can monitor the performance of an object at different times. If it is possible to make a database backup from 12 o'clock to 2 in the evening, you can set up a baseline to monitor the job at this point. Helps administrators determine the Oracle application environment when a database backup job occurs. At this point, it's a good idea to include time information when defining metrics for baseline names. These recommendations can improve the readability of the metrics for baseline names.

Setting "Hours of the day"

I personally think that this parameter is very critical. If the administrator selects the "Hour of the Day" parameter, the result value of the user metrics baseline is the cumulative value in the first few hours. This is useful when you are testing your database. When a user is designing a database, some automated jobs may be designed during a certain period of time. The database administrator needs to test whether the system alert job will be triggered when these jobs occur. At this point, the database administrator can set this parameter. The number of hours in the day selected by the administrator is set to 12. The database system takes the data between 10 o'clock in the morning and 11 as the basis for measurement. If the average metric in this one-hour period exceeds the user-defined percentage, the system alerts you. Therefore, the flexible setting of this parameter can help the administrator to understand the database performance of a specific time period or a particular job. The performance of Oracle subroutines, such as the trigger, is very useful for the impact of the database.

Set the percentage or severity percentage of the warning

Let the system not reach the limit before the alarm or exceeded a certain proportion and then the police. This requires the database administrator to set this parameter according to the actual situation. such as a user's table space capacity may be 10G. When the database controller monitors this object and sets the baseline of the object to 10G, the administrator certainly does not want to call the police when the two has reached 10G. At this point, the database system has been due to the maximum size of the table space has occurred data loss. The administrator wants to alert the administrator when the amount of tablespace usage reaches 90%. Allow administrators to intervene early to prevent accidental losses. To do this, you can define the lower and upper values of the measurement data. In a metric datasheet, there are two text boxes with lower and upper values. After you fill in these two text boxes, the system can use these percentages to calculate the recommended metric valves based on the measure baseline values. Note that you need to enter a percentage value here. If you want to set below the baseline value of 20%, when the alarm occurs, enter 20. Do not enter a percent semicolon.

This percentage is the most skillful place to measure baseline value setup. Because it is directly related to when a certain alarm operation is triggered. This value is set to how much more appropriate, often the public said that the rational, the woman said that the woman is reasonable, there is no standard answer. The author thinks, this needs to adjust according to the actual situation of the enterprise. For example, a database administrator can initially set a lower baseline threshold and then adjust it progressively based on the effectiveness of the alert event. This adjustment process may take half a year or even longer. While this process is relatively lengthy, it is also worthwhile compared to the benefits that automated monitoring brings to database administrators.