Chat now with support
Chat with Support

Foglight for Siebel 5.2.1 - User and Reference Guide

Introducing Foglight for Siebel Systems Foglight for Siebel Systems Summary Dashboard About the Foglight SiebelServer Agent Foglight for Siebel Systems Reference

Sieb_Srvr_Overview graph view

Purpose
When a Siebel task completes its operation, task-level statistics that are gathered dynamically during the operation of a task roll up to the component level. Statistics for all enabled components assigned to a Siebel server are rolled up to the Siebel server level. Siebel stores server statistics as cumulative statistics, that is, since the Siebel server started. These statistics can be used to monitor the progress and performance of a Siebel server.
The Sieb_Srvr_Overview graph correlates the Siebel server’s cumulative response time with its cumulative utilization and throughput.
Figure 10. Siebel Server Overview Since Startup
Use the Siebel Server Overview graph as a progress indicator for the individual Siebel server.
Description of the view
 
Table 15. Sieb_Srvr_Overview graph information
Data Displayed
Average Response Time (min). Average object manager response time, in minutes, since the Siebel server started.
 
Total SQL Execute Time (min). Total CPU time, in minutes, for all the server’s tasks since the Siebel server started.
 
Total CPU Time (min). Total time, in minutes, that the Siebel server spent performing SQL execute operations since the server started. For SELECT statements, this is the amount of time spent identifying the rows if the statement requires a sort or uses FOR UPDATE. For INSERT, UPDATE, and DELETE statements, this is the amount of time spent modifying the data.
 
Total Tasks Started (count). Total number of Siebel tasks started since the Siebel server started.
Where to go next
Drill down on Average Response Time (min) bar, Total SQL Execute Time (min) bar, Total CPU Times (min) bar, or Tasks Started bar. For more information, see Sieb_Srvr_Detail graph view. This displays the correlation of the Siebel server's response time with its utilization and throughput for each refresh interval.

Sieb_Srvr_SQLTime graph view

Purpose
When a Siebel task completes its operation, task-level statistics that are gathered dynamically during the operation of a task roll up to the component level. Statistics for all enabled components assigned to a Siebel server are rolled up to the Siebel server level. Siebel server statistics are cumulative statistics, that is, since the Siebel server started. Among the many statistics that are gathered and rolled up to the Siebel server level are statistics on the time that the Siebel server spends parsing, executing, and fetching SQL statements.
Figure 11. Siebel Server SQL Time Since Last Collection
The Foglight SiebelServer Agent collects delta SQL parse, SQL execute, and SQL fetch statistics for the Siebel server. Foglight calculates the deltas by determining the change in the individual statistic since the last collection. The delta value is stored in the Foglight repository. This allows the Siebel administrator to identify peak SQL usage time periods.
This graph correlates the following Siebel server statistics for each refresh interval:
•
SQL parse time since the last collection.
•
SQL execute time since the last collection.
•
SQL fetch time since the last collection.
Use the Sieb_Srvr_SQLTime graph to analyze how efficiently the Siebel server is interfacing with the Siebel database.
A low SQL parse time indicates a well-designed application that is parsing the SQL statement once but executing it multiple times. Since Siebel dynamically generates SQL statements and uses bind variables, the Siebel administrator is limited in being able to positively affect the SQL parse time. However, if the SQL parse time is high and the Siebel database is Oracle, there are two initialization parameters that can be adjusted to decrease the SQL parse time:
•
SESSION_CACHED_CURSORS. Set this Oracle parameter to a value between 20 and 50. This causes Oracle to ‘remember’ previously parsed SQL and if the user session requests a parse of a SQL statement in the cache, then Oracle avoids the parse.
•
CURSOR_SHARING. Set this Oracle parameter to FORCE. This causes Oracle to substitute literals in SQL statement with bind variables, which allows an otherwise unique SQL to be reused.
High SQL execute time in relation to a low SQL fetch time indicates that more time is being spent performing INSERT, UPDATE, and DELETE statements than SELECT statements.
Low SQL execute time in relation to a high SQL fetch time indicates that more time is being spent performing SELECT statements than INSERT, UPDATE and DELETE statements.
Description of the view
 
Table 16. Sieb_Srvr_SQLTime graph information
Data Displayed
SQL Parse Time (months). The number of months spent translating the SQL statements into an execution plan. This time includes checking for proper security authorization and checks for the existence of tables, columns, and other referenced objects.
 
SQL Execute Time (months). For SELECT statements, the amount of time spent identifying the rows if the statement requires a sort or uses FOR UPDATE. For INSERT, UPDATE, and DELETE statements, the amount of time spent modifying the data.
 
SQL Fetch Time (months). For SELECT statements only, the amount of time spent retrieving the rows that the query returns.
Where to go next
Drill down on any line. For more information, see Sieb_Comp_SQLTime graph view. This display is used to analyze the SQL execute, parse, and fetch to identify which Siebel components are contributing the most to the server’s SQL times.

Sieb_Srvr_Utilization_Overview graph view

Purpose
The Siebel eBusiness application environment is typically made up of multiple Siebel servers with each Siebel server being assigned multiple Siebel components. It is important that all the Siebel servers that make up the Siebel enterprise share the load, and one or more Siebel servers are not overused while others in the enterprise remain underused.
Figure 12. Siebel Server Utilization snapshot
The Foglight SiebelServer Agent collects statistics to enable the Siebel administrator to correlate the Siebel server’s response time, utilization, and throughput.
High CPU consumption alone does not indicate that the Siebel server is overloaded. The server is designed to use all the CPU resources that are available no matter the number of running tasks or object manager sessions. However, excessively high memory utilization by the Siebel server can cause the Siebel server to crash, or limit the number of users who can be serviced by the object managers assigned to the Siebel server.
Is the average response time acceptable? Does it stay within the acceptable range as the number of tasks and object manager sessions increase? As long as the average response time remains acceptable, the Siebel server is not overloaded. It becomes overloaded when its ability to service the running tasks and object manager sessions with an acceptable average response time starts to decline. A sustained decline in the average response time signals an overloaded Siebel server.
Use the Sieb_Srvr_Utilization_Overview graph to determine if the Siebel server is overloaded by correlating the average response time with the number of tasks or object manager sessions the Siebel server is running.
Description of the view
 
Table 17. Sieb_Srvr_Utilization_Overview graph information
Data Displayed
Average Response Time. Siebel server's average object manager response time.
 
Running Tasks. Total number of tasks the Siebel server is running.
 
Interactive Tasks. Number of interactive sessions the Siebel server is running. This includes object manager sessions and other interactive tasks such as the Synchronization Manager and the Server Manager.
 
CPU Time %. Total amount of operating system CPU time the Siebel server is consuming
 
Memory %. Total amount of operating system memory the Siebel server is consuming
Where to go next
Drill down on:
 
•
One of the bars: Average Response Time (min), Total SQL Execute Time (min), Total CPU Times (min), or Tasks Started. For more information, see Sieb_Srvr_Detail graph view. This displays the correlation of the Siebel server's response time with its utilization and throughput for each refresh interval.
 
•
CPU Time % line or Memory % line. For more information, see Sieb_Proc_Utilization graph view. This display is used to analyze the resource utilization for the individual operating system process that make up the Siebel server.
 
•
Avg Response Time line. For more information, see Sieb_Srvr_Detail graph view. This display is used to analyze the Siebel server response time, resource utilization, and throughput since the last collection.
 
•
RunningTasks line or Interactive Tasks line. For more information, see Sieb_Comp_Tasks graph view. This display is used to analyze the number of running tasks, max tasks, running MTServers and max MTServers for each enabled component.

Sieb_Task_Detail Table view

Purpose
The Foglight SiebelServer Agent collects status data for each background and batch Siebel tasks that are currently running, and all Siebel tasks that exited with errors.
Figure 13. Siebel Task Detail table
Use the Siebel Task Detail table to identify the Siebel tasks that are currently running or exited with errors at a specified time period.
If the Siebel administrator identifies a peak usage time period for the Siebel server or an individual Siebel component, use this table view to identify all of the currently running background or batch tasks, or tasks that exit during that peak usage time period.
If a Siebel task exits with an error, examine the task's log file to determine what error caused the Siebel task to exit.
If multiple batch Siebel tasks are running during the peak usage time period and the Siebel administrator can change when the task is started, the Siebel administrator should consider distributing the batch tasks across non-peak usage time periods. Another option is to reassign the Siebel task’s corresponding component to a Siebel server with more available resources, using the Siebel Server Manager.
Description of the view
 
Table 18. Sieb_Srvr_Task Detail Table information
Data Displayed
End Time. The time that a server is shut down. This value is only available when the server is shut down within the Siebel Server Manager.
 
Component Alias. Name of the Siebel component alias being monitored.
 
Task ID. For SELECT statements only, the amount of time spent retrieving the rows that the query returns.
 
Task OS PID. Operating System process id (PID) that is running this task.
 
Task Run Mode. Task’s run mode. For example, Batch, Background, Interactive.
 
Task Start Time. Time the task started.
 
Status. Current run state for the Siebel task. For example, Running, Paused, Stopping, etc.
Where to go next
n/a.
Related Documents

The document was helpful.

Select Rating

I easily found the information I needed.

Select Rating