SAP on IBM i – Update week 02, 2016: Analyzing Resource Consumption on IBM i via DBA Cockpit
In today’s blog post, we’d like to inform you about our latest addition to the DBA Cockpit screens – a monitor for objects and jobs approaching database and operating system limits on IBM i.
For the most important system resources, the IBM i operating system tracks the highest consumption and consumers. This happens automatically as resources are consumed, there is no setup or request needed by the user. The collected information enables users to take action based on trends or limits before a critical situation arises. System limit tracking was introduced with IBM i 6.1 Technology Refresh 6 and is getting enhanced from time to time through newer technology refreshes to track additional limits.
The new screen has been added to SAP NetWeaver 7.0 Enhancement Package 2, SAP NetWeaver 7.3, and higher. Refer to SAP Note 2244176 – IBM i: Enhancements in the DBA Cockpit for IBM i – System Limits about how to get the new option.
To display the new screen System Limits, call transaction DBACOCKPIT and select the system you want to monitor. In the navigation frame, open the folder Diagnostics -> System Limits.
The tracked limits are displayed as a hierarchical list, showing the highest consumption for each limit. The data is filtered to show only information relevant to the monitored SAP system. To offer a quick overview of the system’s current state, the limits are sorted by how close they are to being reached. The hierarchy nodes can be expanded to show more objects pertaining to the respective limit and the historical progression of the resource consumption of each object.
Besides information about the consumer, the level of resource consumption, and the recording date, the rows contain more data depending on the kind of limit. For instance, database object limits contain the long and short names of the objects and the type of object. If the job is still active, you will find more information about the job’s status and its current consumption of system resources like CPU or temporary storage, and which SQL statement it is currently executing.
The screen offers options to sort the displayed rows by different criteria, to group the limits, and to restrict the amount of displayed data. After you have finished specifying all query options, select the pushbutton Apply Selection on the upper left to refresh the display according to the selected options. For better performance, this will adapt the display of the already selected data. For reading the newest data from the database, select the pushbutton Refresh instead.
Please note that the most recent record of any resource consumption may not be identical to the current level of resource consumption for this object. In order to keep the impact on the operating system low, system limit tracking periodically saves a new record depending on the progression of the consumption. Changes to an object between these save points are not logged.
This means that the System Limits screen can point you towards potential resource bottlenecks, but you still need to analyze the actual situation by checking the IBM i operating system or DB2 for i database directly.
The system limits tracking also includes an automated pruning of old entries to save database space. As of IBM i 7.1, the pruning behavior can be controlled by customizable parameters. Therefore, if the monitored SAP system runs on IBM DB2 for i 7.1 or higher, the output table contains an additional pushbutton Pruning Controls for displaying these parameters. If your SAP user has sufficient privileges for SAP authorization object S_RZL_ADM, you may also edit them here.
Please note that these parameters are DB2 for i supplied global variables. Thus, any changes will affect all systems on the same LPAR. Please also note that changed values only become effective after the next IPL.
It should be worthwhile for every system administrator to check this at least once because most customers are completely unaware of whether or not they are close to hitting any limits. If you find that you are not close to any limits you might not need to check this very often. If you find that some limits are reachable, then you should check how quickly the system is approaching them by looking at the data over time. This can help to determine by when you need to take some preventative action.