[Gluster-devel] I/O fair share to avoid I/O bottlenecks on small clsuters

[gordan at bobich.net]

On Sat, 30 Jan 2010, Ran wrote:

> Hi ,
> I recently posted an issue regarding a situation that happen when
> say 1 virtual machine images takes down the entire server IO
> then the entire storage become slow in such way that noting work
> emails , web etc...
> gluster gays posted that virtual machine is a main goal of the storage
> and that they probebly implement a fair share IO options to avoid this cases .
>
> Can anyone tell me what are the plans for this , it apear to me that
> this is 1 of the most important issues on such storage seens it is not
> possible to run more then a few virtual machins in parallel .

Unless I am misunderstanding what you are saying, the question is 
rather non-sensical. The whole point of virtualization is that you can 
ration the resources between VMs. It is the duty of the hypervisor to 
implement the resource restrictions and fair resource sharing. The VM 
_should_ utilize all the resources that the hypervisor lets it.

In other words, if you have a VM that is consuming all the resources of 
the server, that means you either have a lame hypervisor or that you 
haven't configured it properly. A VM should _NOT_ be expected to 
self-censor, and certainly not on application level.

Gordan

P.S.
I'm not on the users list, so can't forward this there, feel free to 
bounce your reply there with comments in it, though.