Public discussion of the formula for calculating the cost of resources in the MQL5 Cloud Network - page 46

 
Andrey Khatimlianskii:

Are there really any brakes? I haven't opted with GAs in the cloud for a long time.

A year ago I optimised in the cloud, there were hangs on individual passes. I think it was because of weak machines. Sometimes an error occurred due to insufficient memory. Now I am preparing a new project and would like to prevent this.
For example, if you have 8 cores, it is desirable to have at least 16 Gg of RAM.
 
elibrarius:
A year ago I optimised in the cloud, there were hang-ups of individual passes. I think it was because of weak machines. Sometimes an error occurred due to lack of memory. Now I am preparing a new project and would like to prevent this.
For example, if you have 8 cores, it is desirable to have at least 16 Gg of RAM.

Yes, it makes sense.

 

Why does the PR drop after passes, and why is it different for all agents?

AgentsPR

 
Dmitry Melnichenko:

Why does the PR drop after passes, and why is it different for all agents?


Your number of passes is different - the PR calculation is not averaged. And obviously performance per thread drops when you activate 4 threads, since you actually have 2 cores, not 4 processor cores.

 
Aleksey Vyazmikin:

Your number of passes is different - the PR calculation is not averaged. And, clearly the performance per thread drops when you activate 4 threads since you actually have 2 cores, not 4 on the processor.

thanks
 
Can you tell me where I can find a cost estimate for Cloud testing?