Geek Answers Handbook

Why does .NET use so much CPU, is this a problem?

(This was supposed to be a general hypothetical question, I squealed that .NET was a pig and begging for reasons. Actually, this was not a question about my specific application.)

I am currently rewriting old C ++ code in C #. We migrate all outdated applications. I have C ++ applications that use MAX 3% CPU. Mostly they do not use them. Then I take the code, copy and paste, and then reformat the C # syntax and the .NET and BAM libraries! 50% processor. What is the reason for this? At first I thought it was JIT, but even after each code path was an exercise, and it was all JIT ed, the same problem.

I also noticed that huge memory is increasing. Applications that run a full 9 MB download now start at 10 MB and run at 50 MB. I understand that equipment is cheap, but I want to understand what causes this. Is this cause for alarm, or is it just such a pig?

Update 1 Reply to Skeet

I am familiar with C #. I change things to Linq and so on. I usually take the code and reduce the number of lines, etc. Could you give some more examples of what C ++ people did wrong in .NET?

Update 2

This should have been a general question, but the specific application that has the problem is as follows.

It has a stream that uses the ODBC driver to retrieve data from the db paradox. He then uses Linq to convert it to SQL db and publish it. I run it through the ANTS profiler, and it seems that filling the data set takes the most time. Then comes the Linq wiring. I know that some of my areas are using reflection, but I don’t see how to do what I need to exclude it. I plan to change my string to string collectors. Is there a difference between the two?

(int)datarow["Index"]

and

ConvertTo.Int32(datarow["Index"])

I changed all string concatenation to format strings. It has not diminished on the head. Does anyone know the difference between a data reader and data adapter and datasets?

+5
5

# .NET? , ++, , . . (, , .) , ++, .NET-, , , .

, , , , . ? 2, , , ...

.NET, , , ++-, , , . 50 9 , , , .

, . JetBrains dotTrace Profiler, .

+17

AFAIK (int)datarow["Index"] ConvertTo.Int32(datarow["Index"]). , :

int orderIndex = <order of Index column in projection list>;
using (OdbcDataReader rdr = cmd.ExecuteReader(CommandBehavior.SequentialAccess))
{
  int Index = rdr.GetInt32(orderIndex);
}

SeqentialAccess - SQL, , .

, , , . , , "". ( , Rushmore). 1 , (, DataTables).

, CLR . IL NGEN. CLR, , . GC pre-emption "pause" , ( , GC, , ). CLR , , , , TLB L2.

, , "++ #" - - , , C ++ ( " ", ), , C.

+6

. 3% 50% , ?

- . 30 50 .NET - . , .NET , , ( , ), , .

+1

, , , , .

, , .

0

:

1) . ++ . , ( ) .

"" , . , , . , , , , . memo9ry, gc inv_ked.

2) Is the program running at the same time or less? If it uses 5 times more processor power, but ends in 1/5 of the time, then the CPU used is basically equivalent.

3) Yes, .NET is probably a pig

0
source

More articles:


All Articles