Open Stack

On Tue, 07 Jun 2016 17:08:09 +0200, Pranav Salunke wrote:
> I stumbled upon a very rare and interesting case (logs attached) while
> deploying the training-labs cluster with KVM backend (using archlinux).
> 
> http://paste.openstack.org/show/508650/
> 
> My processor was quite heavily overloaded since I was using the laptop
> (quite powerful) to do some deeplearning training batches in parallel.

Are you sure that's the right paste? There's the error you will get if
your /tmp directory is not big enough to hold hold a temporary copy
of the basedisk (used for compression). I don't see anything rare or
particularly interesting there.

Admittedly, it would be nice if we handled this condition more gracefully.

Roger