trchandra
27th April 2002, 01:48
Hi
How to convert ticks into time (seconds) that the program is using in baan?
When u open Opiton Dialog and goto Shell...
You can issue command 'ps' to see current processes...
there one of the column is TICKS...
Can we get time from this ticks?
appreciate your help
Bogdan
29th April 2002, 15:27
Hi,
Well, acording to baan help TICKS=number of used CPU ticks;
Based on what I know about CPU's that mean the number of CPU clock units the CPU need to run a specific command(s) (I mean assembler commands). This mean that this number depends by CPU speed (not the "comercial" one but the real one-you will need technical specifications from the manufacturer) and let's call it work flow of the CPU (it's not necessary to be the same number for a specific process; what I mean is that there could be different depending by the CPU load, not much, only a few tick -because some assembler commands can be executed in x or y ticks).
So, for accuracy you will need to know the MACHINE CODE of the process and then start to count for each command the number of ticks; this is if you want to precalculate this time.
If you just need the value simply multiply the number of ticks with internal clock speed of the CPU.
Hope this will help.
Bogdan
NPRao
5th March 2003, 00:11
Ravi,
Please refer to - Application Performance Guide - M2017 B US.pdf - Chapter 6- Profiling and Tracing.
Profiling and tracing
BAAN Application Performance Guide
53
6.2.3 Analyzing the profiler output
The meaning of the column headers is:
Count Number of times the function is carried out
Utime User time in seconds
Call Average time for certain function call
Perc % Percentage of total time
Stime System time in seconds
Rtime Real time in seconds
Alloc Number of allocated bytes
Free Released memory
Ticks Number of bshell ticks
Function Name of executed function
Each time in the output is the time used only in that function. The profile figures are
not cumulated.
So, if another function B has been called in function A, the time used in function B is
not taken into account for the time logged in function A.
The Rtime column is the most important column.
This is the time a function really takes (a user must wait).
For this reason, it is advisable to search for the highest values in this column.
The Count column is also very important.
The numbers in the Count column can be verified to see that the number is not too
high.
Because: count * call = rtime.
The Ticks column is also important.
The number of ticks indicates how may CPU ticks (instructions) are handled in the
bshell.
There is no relation between Rtime and Ticks because one tick take a fraction of a
second, but it can also take many minutes.
For example: A full table scan can be carried out in one bshell tick and can take
several seconds (or minutes). A calculation can be carried out where thousands of
bshell ticks are needed and only one second or less is taken.
So, a function with a high rtime and a low number of ticks, is probably a function in
which intensive database action has been implemented.
A function with a low rtime and a high number of ticks, is probably a function in
which a calculation or a fast database action with a lot of fetches has been done.