Errors, bugs, questions - page 319

 
a483483:

This is a screenshot of the screen after the strategy test is stopped. You can see that several processes are holding 4GB of memory each.

Is this a bug?

Wow 100 gigabytes of RAM!
The tester keeps the caches raised for a while for a quick test after a short time on the same history interval...
 
a483483:

This is a screenshot of the screen after the strategy test is stopped. You can see that several processes are holding 4GB of memory each.

Is this a bug?

Is the terminal itself a 404 build or older?

If 404, then it is a bug. Can you attach agent side logs?

 
a483483:

This is a screenshot of the screen after the strategy test is stopped. You can see that several processes are holding 4GB of memory each.

Is this a bug?

Local agents hang in memory with their caches raised for another 5 minutes after stopping the test, waiting for new tasks. If no one loads them in 5 minutes, they are automatically unloaded and all resources are freed.

This waiting mechanism allowed agents to increase response speed on repeated tasks at the expense of ready/raised infrastructures.

ps: these are local agents. remote agents running as services are not offloaded, but also release resources after some downtime.

 

Thanks for the replies about the 2 e-mails.

I just came across an interesting behaviour in the strategy tester (MT5):

Stops

How can that be? I'm just wondering...

 

Terminal version 5.0 404 (Fibo)

Windows 7 x64 Home extended

Dear Sirs! This is the 3rd time I have addressed this problem. Creating an external library under Windows7 x64 leads to some miracles:

Once again, when you take it to a degree, the result is infinity! The strangest thing: at first I created the library and tested it in a separate indicator, everything was OK. Then I updated the functions and in this case...

I tried to play games like _Digits replace with a variable, it didn't help - the result is still MathPow(10,_Digits)=INfinity!

Other functions also start to freak out.

This problem is not present if the functions are moved to the body of the program, and on Windows XP this problem simply does not exist!!!

Thanks.

 
The stoploss is 2 pips, which is apparently not enough in a working account. MathPow will be sorted out, but it's better to provide the code that generates the error right away. Without the code you can write many times with no effect.
 
Renat:
The stoploss is 2 pips, which is apparently not enough in a working account. MathPow will be sorted out, but it's better to provide the code that gives you the error right away. Without the code, you can send it many times without any effect.
The code and the result of the code repeatedly sent TO ALEXVD, as well as via Timvox...
 
Thank you, we will check and correct.
 
Renat:
Thank you, we will check and fix it.

Renat!

I asked a question once, but everyone ignored it.

Why are there so many errors in the x64 version? Is the code fundamentally different there or what is it related to?

I've just never encountered x64 compilers

 
AlexSTAL:

Why is there so many errors in the x64 version? Is the code fundamentally different there or what is it related to?

I've just never encountered x64 compilers

The code is fundamentally different