Currently there is no option to diagnose performance issues without enabling tracing.
Customers need to identify where performance bottlenecks are without enabling tracing and without affecting overall performance
CA SSO / SiteMinder often gets the blame for failures in back end database or network. These types of problems need to be identified quickly and accurately. Hence we added few more things to smpolicysrv stats to collect the data for the analysis and removed unwanted things.
New Stats Available:
Cumulative Number of Requests
Average Transaction Time (millisec)
Average Throughput (req/sec)
Average Wait Time in Queue (millisec)
6241/2681244560][Wed Feb 03 2016 15:03:07][CServer.cpp:4729][INFO][sm-Server-02030] Thread pool: Msgs=58844 Throughput=238.235/sec Response Time=4.820ms Wait Time In Queue=16.886ms Max HP Msg=1 Max NP Msg=50 Current Depth=0 Max Depth=50 Current High Depth=0 Current Norm Depth=0 Current Threads=8 Max Threads=8 Busy Threads=0
Msgs now correctly measures total number of requests and not number of poll attempts.
Average Wait Time in Queue signifies time spent between ‘Enqueue’ and ‘Dequeue’.
Avg Transaction Time is not the inverse of Throughput. To get Throughput we use total run time of test. To get Avg Transaction Time, we use only time spent processing transactions.
‘Waits’ and ‘Misses’ did not reflect a meaningful data-point and have been removed.