| Measurement |
Description |
Measurement Unit |
Interpretation |
| free |
Indicates the number of work processes of this type that are currently free. |
Number |
A high value is desired for these measures. Sufficient number of free work processes are needed to handle sudden spikes in activity. It is important that free dialog work processes are available so that administrators will be able to login and troubleshoot other issues. |
| pctfreewp |
Indicates the percentage of work processes of this type that are currently free. |
Percent |
| onhold |
Indicates the number of work processes of this type that are currently on hold. |
Number |
Work processes are held when a particular user holds the work proceses while waiting for some message. These processes cannot be released to the free pool until the message is received. Work processes on PRIV mode are also counted as on hold. Some issues can cause work processes to be held indefinitely thereby severely limiting/exhausting the number of free work processes.
The detailed diagnosis of this measure if enabled, lists the process ID, CPU utilization, user, report being run, table being accessed, elapsed time etc., |
| pcthold |
Indicates the percentage of work processes of this type that are currently on hold. |
Percent |
A low value is desired for this measure. If there is a sudden surge in the value of this measure, then it would adversely affect the performance of the server by exhausting the number of free work processes. Using the onhold measure administrators may figure out the work processes that are held for a longer duration and examine the real cause. |
| privmode |
Indicates the number of work processes of this type that are currently in PRIV mode. |
Number |
A work process enters PRIV mode when either extended memory is exhausted or the process has used the amount of memory specified in ztta/roll_extension. The work process then starts utilizing the heap area which is not appreciated. If the memory utilized is greater than the one set against the abap/heaplimit profile parameter, then the work process has to be restarted after terminating the user (i.e., after user sign out process). This will help administrators to restore the work processes from the PRIV mode.
A low value is desired for this measure. An increasing trend in the value of this measure indicates severe performance bottleneck.
The detailed diagnosis of this measure if enabled, lists the process ID, user, report being run, table being accessed, elapsed time etc., |
| pctpriv |
Indicates the percentage of work processes of this type that are currently in PRIV mode. |
Percent |
A low value is desired for this measure. The system will automatically terminate the executing transactions if more than a certain percentage of work processes are in PRIV mode and the user is inactive for a certain period of time. |
| running |
Indicates the number of work processes of this type that are currently running. |
Number |
Comparing the value of this measure across the types helps you in identifying the work processes type that is highly utilized. This data can be very useful for capacity planning and fine tuning.
The detailed diagnosis of this measure lists the work process details thereby providing a snapshot view of the work processes that are currently executing. |
| stopped |
Indicates the work processes of this type that are currently in Stopped state. |
Number |
The work processes would be in Stopped state when the work processes have been aborted and are not to be restarted. This usually happens because of a serious startup issue for the work process which needs to be fixed before the work process can be started again.
The detailed diagnosis of this measure lists the details about these work processes such as the process ID, user, report being run, table being accessed, elapsed time etc., |
| total |
Indicates the total number of work processes of this type that are configured to run. |
Number |
Comparing the value of this measure helps administrators in planning the capacity of each work processes type and fine tune it accordingly. |
| dumps |
Indicates the total number of created by the work processes of this type during the last measurement period. |
Number |
The detailed diagnosis of this measure if enabled, lists the name of the work processes that had errors / dumps, the CPU utilization, number of dumps thrown, user, report running, table accessed etc., Using these details administrators can deduce the reason behind the creation of dumps. |
| requestqueueutil |
Indicates the percent of request queue for the work processes of this type that is currently utilized. |
Percent |
A high value for this measure indicates a sudden surge in the user activity which helps administrators to validate the distribution of work processes. |
| cpuutil |
Indicates the overall CPU utilization of the work processes of this type. |
Percent |
This measure indicates service level CPU utilization of the various services (work process types) such as Dialog service, Update service, Spool service etc.,
This mesure helps administrators to identify the resource intensive work processes and figure out the real reason that is hogging the resources.
The detailed diagnosis of this measure if enabled, provides a sorted list of work processes by CPU utilization. |
| waitingreq |
Indicates the number of requests for this work process type that are waiting to be processed. |
Number |
A higher number of waiting requests results in high dispatcher wait time and response time. This helps to provision and distribute work processes appropriately. |