← Back to team overview

openstack team mailing list archive

Re: Specification of nova-diagnostics


I'm not familiar with nova-diagnostics, but we use collectd to gather
information about each virtual machine running in KVM. We get information
about cpu usage, memory, disk io, and network io per instance. Obviously
it's not nicely integrated with nova commands, but it works in our
deployments. And it comes in the distro by default.

Let me know if I can help you in this blueprint: the only reason now in our
distro to use collectd is kvm instances performance. If we can get from
nova we should go that way in the future.

Diego Parrilla
*www.stackops.com | * diego.parrilla@xxxxxxxxxxxx** | +34 649 94 43 29 |
* <http://www.stackops.com/>


On Mon, Mar 26, 2012 at 6:06 PM, Leander Bessa <leanderbb@xxxxxxxxx> wrote:

> Hello,
> I've been trying to implement the nova-diagnostics command  for libvirt.
> I'm using the devstack installation on ubuntu 12.04. So far i've managed to
> gather information regarding CPU, IO (network and disks). The memory
> reading only works with the virsh command and does not report the actual
> memory usage.  As pointed out to me previously, i've been looking at the
> implementation of nova-diagnostics for the xenapi in order to try and
> figure out how to supply the data back to openstack. However, i can not
> make any sense of it. All i know is that it returns a set of keys with
> values, which i suppose represent the values of the readings. What i wish
> to know is if there's any document or specification which i can read which
> defines what nova-diagnostics should return and under what format.
> Regards,
> Leander
> _______________________________________________
> Mailing list: https://launchpad.net/~openstack
> Post to     : openstack@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~openstack
> More help   : https://help.launchpad.net/ListHelp

Follow ups