Metatester io / Harddisk 100%

New comment
 

Hi there, 

I noticed, that on some mashines the Metatester consumes a lot harddisk... in the windows ressource monitor the harddisk is used with 100% ( running 10 tasks f.e.), therefore the process is more or less stalled ...

this only happens on some mashines.. the RAM is not used at all , so there is no reason for swapping... 

In the agent logs I can not see anything spectacular ... 


Any ideas?

 
Filip:

Hi there, 

I noticed, that on some mashines the Metatester consumes a lot harddisk... in the windows ressource monitor the harddisk is used with 100% ( running 10 tasks f.e.), therefore the process is more or less stalled ...

this only happens on some mashines.. the RAM is not used at all , so there is no reason for swapping... 

In the agent logs I can not see anything spectacular ... 


Any ideas?

Heavy or not heavy depends on your hardware spec. Try SSD, OHLC mode and reduce the number of agents/cores .
 
Filip:

Hi there, 

I noticed, that on some mashines the Metatester consumes a lot harddisk... in the windows ressource monitor the harddisk is used with 100% ( running 10 tasks f.e.), therefore the process is more or less stalled ...

this only happens on some mashines.. the RAM is not used at all , so there is no reason for swapping... 

In the agent logs I can not see anything spectacular ... 


Any ideas?

From where did you get which numbers? Did you do your tests/optimizations or were they from the MQ cloud?
For a chance to get help you have to provide detailed information like logs, screenshots, ticks or no ticks, test duration, ...!!

 
Filip:

Hi there, 

I noticed, that on some mashines the Metatester consumes a lot harddisk... in the windows ressource monitor the harddisk is used with 100% ( running 10 tasks f.e.), therefore the process is more or less stalled ...

this only happens on some mashines.. the RAM is not used at all , so there is no reason for swapping... 

In the agent logs I can not see anything spectacular ... 


Any ideas?

i think you need to do a few more tests. if you did only a few tests, you would note that at start of EVERY test, whether you use 1 core or 10, even a ssd is often at 100% for the first few minutes before dropping off.

 
Filip:

Hi there, 

I noticed, that on some mashines the Metatester consumes a lot harddisk... in the windows ressource monitor the harddisk is used with 100% ( running 10 tasks f.e.), therefore the process is more or less stalled ...

this only happens on some mashines.. the RAM is not used at all , so there is no reason for swapping... 

In the agent logs I can not see anything spectacular ... 


Any ideas?

At the risk of me sounding like a broken record...

If you don't want to pay for Cloud Agents to speed up the process of backtesting, limit every backtest to one Input and one less Steps than your total number of cores as listed on the Agents tab of the Tester.

 
Filip:

Hi there, 

I noticed, that on some mashines the Metatester consumes a lot harddisk... in the windows ressource monitor the harddisk is used with 100% ( running 10 tasks f.e.), therefore the process is more or less stalled ...



HDD or SSD?  Use a SSD, those with ram cache one is better.

New comment