Hello,

It is nothing specific to JDE in regards the performance. You should always have in your mind general principles of software programing. For example when fetching data, in case you want to fetch data having specific condition, then you should use that condition to fetch the data and not walk through a full data set.

Timing for execution of a line of code cannot be measured.
I believe you are referring to the Event Rules timings. Even if you are referring to NER or BSFN lines, please keep in mind that an instruction in C, at compilation is converted to assembly language and different instructions in C can be implemented in assembly in several commands. Each assembler line takes one CPU tick to execute. The higher CPU frequency, the quicker multiple assembler lines (codes) are executed. This is a trivial explanation and is not involving low level analysis of code execution.

You may start design an UBE which will make use of NER and BSFN and check the performance. You always can see where the delays are occurring if you will run the UBE with debug logging on and then use the Performance Workbench (for example) to check the timings and how many times data has been fetched from a table.

Please check also the following KM Documents:
- JD Edwards EnterpriseOne Diagnostic Tools: Performance Workbench (Doc ID 2013059.1)
- E1: ENV: Performance Workbench Tool Information for Analyzing JDEDEBUG Logs (Doc ID 654972.1)

Thank you!

Best regards,
Cezar