ST03N: Workload Monitor [Monitoring]
ST03N is used to analyze statistical data for the ABAP
kernel and monitor the performance of a system. You can display the
total values for all instances, and compare the performance of
particular instances over a period of time.
The workload overview provides
system administrators with various detailed information about the most
important workload data, such as the CPU time, the number of database
changes, the response times, and so on. You can display the workload
overview for all task types (Dialog, Background, RFC, ALE,
and Update), or only for one particular task type.
Workload Overview :-
Processing time – This is equivalent to response time minus the sum of wait
time, database request time, load time, roll time, and enqueue time
Hint: > 2x of CPU time
Probs: Hardware
Hint: > 2x of CPU time
Probs: Hardware
CPU time – A work process uses the CPU.
Hint: 40% of response time]
Probs: CPU bottleneck
Solution:
In ST06,
Go to Detail Analysis Menu -> Top CPU , check existence of Non-SAP (external) programs by sorting by CPU time. Try to run these external programs in offline hours.
In ST02,
Check for any swapping happening in all the buffers. If there is high swapping for any buffer, increase the size of the buffer.
Hint: 40% of response time]
Probs: CPU bottleneck
Solution:
In ST06,
Go to Detail Analysis Menu -> Top CPU , check existence of Non-SAP (external) programs by sorting by CPU time. Try to run these external programs in offline hours.
In ST02,
Check for any swapping happening in all the buffers. If there is high swapping for any buffer, increase the size of the buffer.
Response time – The time when a dialog process sends a request to a dispatcher work process, and the dialog is complete and the data is transferred to the presentation layer. The response time does not include the time for transferring the data from the SAP front end to the application server.
Hint: 1 second (dialog), <1 second (update)
Wait time – The time when a user request sits in the dispatcher queue. It starts when user request is entered in the dispatcher queue; and ends when the request starts being processed.
Hint: < 10% of response time
Probs: long running tasks, locked tasks, not enough work process
Solution:
In SM50,
Look for all the configured work processes are in Waiting or Running state. If all the wotk processes are running state, then increase the number of Dialog work processes.
In SM66,
This monitor will help to analyse the total work processes configured in all the servers and instances.
Probs: long running tasks, locked tasks, not enough work process
Solution:
In SM50,
Look for all the configured work processes are in Waiting or Running state. If all the wotk processes are running state, then increase the number of Dialog work processes.
In SM66,
This monitor will help to analyse the total work processes configured in all the servers and instances.
DB calls – Number of parsed accesses to the database.
Hint: DB calls/requests good ratio is 1:10 = efficiency table buffering
Hint: DB calls/requests good ratio is 1:10 = efficiency table buffering
DB requests/DB Time – The time when a database request is put through to the
database interface & when the database interface has delivered the result.
Hint: 40% of response time
Probs: CPU/memory bottleneck on DB server, expensive SQL statement, missing indexes, small buffer, missing statistics
Solution:
In ST04,
- Database buffer quality (> 95%), if <, increase database buffer cache size.
- Reads/User Calls (< 30), if >, the expensive SQL statements need to be tune. Some of expensive SQLs statement problems:
i) incorrect index access (Solution = create new index or reorganize the index)
ii) high table size (Solution = archive the old entries)
Hint: 40% of response time
Probs: CPU/memory bottleneck on DB server, expensive SQL statement, missing indexes, small buffer, missing statistics
Solution:
In ST04,
- Database buffer quality (> 95%), if <, increase database buffer cache size.
- Reads/User Calls (< 30), if >, the expensive SQL statements need to be tune. Some of expensive SQLs statement problems:
i) incorrect index access (Solution = create new index or reorganize the index)
ii) high table size (Solution = archive the old entries)
Average load & generation – The time needed to load and generate objects.
Hint: < 10% of response time, < 50ms
Probs: Program buffer, CUA buffer, screen buffer too small
Hint: < 10% of response time, < 50ms
Probs: Program buffer, CUA buffer, screen buffer too small
GUI time – Response time between the dispatcher and the GUI during the roundtrips (roundtrips are communication steps between the SAP system and the front end during a transaction step).
Hint: < 200ms
Probs: network between GUI & SAP]
Solution:
In ST06,
Go to Detail Analysis Menu -> LAN Check by PING. If there is high Avg. time or Loss time for any presentation servers, means there are some settings need to be change for the presentation server.
In SE38,
Execute PROFGEN_CORR_REPORT_5 report. From the output check if any user assigned with > 1000 user menu nodes.
Probs: network between GUI & SAP]
Solution:
In ST06,
Go to Detail Analysis Menu -> LAN Check by PING. If there is high Avg. time or Loss time for any presentation servers, means there are some settings need to be change for the presentation server.
In SE38,
Execute PROFGEN_CORR_REPORT_5 report. From the output check if any user assigned with > 1000 user menu nodes.
Roll in time - The time needed to roll user context information
into the work process.
Hint: < 20ms
Probs: SAP memory configuration (extended memory, roll buffer)
Hint: < 20ms
Probs: SAP memory configuration (extended memory, roll buffer)
Roll wait time – Queue time in the roll area.
Hint: < 200ms
Hint: < 200ms
Probs:
network between GUI & SAP
