OSBot stability with 20+ bots

[Abuse]

Hi,

 

I've been wondering for a while if it is possible to improve the performance/stability of multiple OSBot instances, more particularly stealth injection.

 

I am currently running 4 dedicated servers with each having around 20 clients sitting at 70-90% CPU load, however some clients tend to randomly freeze (Black screen) after a few hours of runtime, this seems to be dependent on the amount of running bots, 10 usually don't cause any issues.

I've tried both Linux and Windows with the same results, however linux struggled with more than 10 clients

 

In terms of resource usage, my dedi boxes run at around 60% memory usage, 512mb dedicated to each instance so that doesn't seem to be the issue

 

- Abuse

Edited July 12, 2016 by Abuse

[Muffins]

It's probably memory leaks caused by scripts running for a long time, you could try increasing the allocation size of each client

[Santonio]

It's probably memory leaks caused by scripts running for a long time, you could try increasing the allocation size of each client

+1

[Aiban]

There is a low resource mode available with CLI try it out

[Token]

Hi,

 

I've been wondering for a while if it is possible to improve the performance/stability of multiple OSBot instances, more particularly stealth injection.

 

I am currently running 4 dedicated servers with each having around 20 clients sitting at 70-90% CPU load, however some clients tend to randomly freeze (Black screen) after a few hours of runtime, this seems to be dependent on the amount of running bots, 10 usually don't cause any issues.

I've tried both Linux and Windows with the same results, however linux struggled with more than 10 clients

 

In terms of resource usage, my dedi boxes run at around 60% memory usage, 512mb dedicated to each instance so that doesn't seem to be the issue

 

- Abuse

Refrain from limiting the memory usage of a client. If it has to go above that limit doesn't mean it will simply stop at 512 mb and won't have any effect. It will actually make the client freeze because you are putting too much pressure on the garbage collector with your memory limit.

[Charlotte]

I run 20+ bots with 300~350mb per bot. All minimized, sitting at about 20-30% cpu usage. Then again if you require more ram, do not limit to just 512.

 

[Harry]

It's probably memory leaks caused by scripts running for a long time, you could try increasing the allocation size of each client

This.

 

you should also contact the Scripter and see if they can improve on the script on using less memory/memory leak issues. It's usually caused by poor coding or bad methods that take up a lot of CPU/memory after a long time.

[Abuse]

All right, thanks for your help all! It seems that the problem was related to my RDP connection, each time i would disconnect/reconnect to the RDP one or more bots would freeze.

If anyone has similar problems, force your remote server to display to a physical display output

1. Right click on the desktop, click 'Screen Resolution'
2. Click 'Detect' button on the next screen (when click on Detect, a new second monitor was shown on the right of the main display)
3. Click 'Another display not detected' and under the multiple displays option select 'Try to connect anyway on: VGA'
4. Click 'Apply'

This seems to do the trick.

 

 

[Kittens]

It's probably memory leaks caused by scripts running for a long time, you could try increasing the allocation size of each client

 

 

do as muffins said, i've only tested it for 2 hours and it seems to work, not 100% sure though. I'm going to be stress testing it hard to get 10+ hour proggys with what muffins has suggested and if it work's I will let you know