03-05-2012, 04:19 AM
Hello there,
I rent a dedicated server (which contradicts the title of this thread ) which uses Windows Server 2008 (32-bit or x86) with the following specifications:
CPU: Intel Q6600 @ 2.4Ghz (1 Processor, 4 Cores)
Memory: 4GB
Storage: 100GB
Network: 1Gbps Port on a 1Gbps line
Currently running on the server are:
- 1 Garry's Mod server (16 slot)
- 1 Counter Strike Source server (24 slot)
- 2 L4D2 servers
- 1 Minecraft (1GB allocated)
However due to the set up my hoster decided to go with to ensure they had the ability to back up my system and recover it when necessary remotely, the machine I connect to is a virtual machine (more specifically VMWare) running on the dedicated hardware which has access to 100% of the hardware resources.
Whilst this seems like a good idea, I am having issues... Whenever something causes additional I/O usage, the CPU goes from 0 - 20% to 90% -100%.
Attached to this post are images of what happens during e.g. a game server restarting. The image named as latency.jpg, shows a latency against time graph of one of our servers which is experiencing the issue caused by another server restarting (The spikes show points where the server ping shot up to between 300 and in some cases even 2000, whilst the large red bar represents me in the process of connecting to the remote desktop), the second image (named remote desktop.jpg) is annotated and shows what is happening on screen whilst a server is restarting).
From what I have read, this could be a fault with VMWare, where additional I/O usage causes server performance degradation.
The issue can occur whenever:
- A Game server restarts
- A Game server changes level
- Our FTP server has to write to the disk with transactions
- Files are copied from one section of the disk to the other
In all circumstances, the server box returns to normal operating
conditions after these processes have been finished/cancelled. I have tried other tests not related to the Hard drive interaction, but they do not produce the same effects.
If you believe I am incorrect about the I/O usage causing it please do enlighten me as to what you believe may be causing it. If you do not believe it to be related to VMWare, then what could be causing this issue in your honest opinion?
Any help greatly appreciated
I rent a dedicated server (which contradicts the title of this thread ) which uses Windows Server 2008 (32-bit or x86) with the following specifications:
CPU: Intel Q6600 @ 2.4Ghz (1 Processor, 4 Cores)
Memory: 4GB
Storage: 100GB
Network: 1Gbps Port on a 1Gbps line
Currently running on the server are:
- 1 Garry's Mod server (16 slot)
- 1 Counter Strike Source server (24 slot)
- 2 L4D2 servers
- 1 Minecraft (1GB allocated)
However due to the set up my hoster decided to go with to ensure they had the ability to back up my system and recover it when necessary remotely, the machine I connect to is a virtual machine (more specifically VMWare) running on the dedicated hardware which has access to 100% of the hardware resources.
Whilst this seems like a good idea, I am having issues... Whenever something causes additional I/O usage, the CPU goes from 0 - 20% to 90% -100%.
Attached to this post are images of what happens during e.g. a game server restarting. The image named as latency.jpg, shows a latency against time graph of one of our servers which is experiencing the issue caused by another server restarting (The spikes show points where the server ping shot up to between 300 and in some cases even 2000, whilst the large red bar represents me in the process of connecting to the remote desktop), the second image (named remote desktop.jpg) is annotated and shows what is happening on screen whilst a server is restarting).
From what I have read, this could be a fault with VMWare, where additional I/O usage causes server performance degradation.
The issue can occur whenever:
- A Game server restarts
- A Game server changes level
- Our FTP server has to write to the disk with transactions
- Files are copied from one section of the disk to the other
In all circumstances, the server box returns to normal operating
conditions after these processes have been finished/cancelled. I have tried other tests not related to the Hard drive interaction, but they do not produce the same effects.
If you believe I am incorrect about the I/O usage causing it please do enlighten me as to what you believe may be causing it. If you do not believe it to be related to VMWare, then what could be causing this issue in your honest opinion?
Any help greatly appreciated