On Fri, Sep 1, 2017 at 10:42 AM, Rua, Philippe (Nokia - FI/Espoo)
<philippe....@nokia.com> wrote:
Will it be possible to include network bandwidth as a resource in
Nova scheduling, for VM placement decision?
I think it will.
Context: in telecommunication applications, the network traffic is an
important dimension of resource usage. For example, it is often
important to distribute "bandwidth-greedy" VMs to different compute
nodes. There were some earlier discussions on this topic, but I could
not find a concrete outcome. [1][2][3]
After some reading, I wonder whether the Custom resource classes can
provide a generic mechanism? [4][5][6]
Here is what I have in mind:
- The VM need is specified in the flavor extra-specs, e.g.
resources:CUSTOM_BANDWIDTH=123.
- The compute node total capacity is specified in host aggregate
metadata, e.g. CUSTOM_BANDWIDTH=999.
I'm not aware of any feature that considers aggregate metadata key as
resource inventory. As far as I know you have to define new resource
providers for your CUSTOM_BANDWIDTH resource via the placement API and
you have to report the 999 as inventory on those resource providers
also via placement API. Also don't forget to connect your resource
provider to the existing compute resource providers via an aggregate
(this is an aggregate in placement which is different from the host
aggregate concept in nova). This review contains some test cases that
can help you how to set things up
https://review.openstack.org/#/c/497399
- Nova then takes care of the rest: scheduling where the free
capacity is sufficient, and performing simple resource usage
accounting (updating the compute node free network bandwidth capacity
as required).
With the above flavor extra spec as request and the above resource
provider setup nova will do the rest of the resource accounting for the
your custom resource. Except in case you hit one of the bugs we
discovered in this area
https://bugs.launchpad.net/nova/+bugs?field.tag=placement
Is the outline above according to current plans?
If not, what would be possible/needed in order to achieve the same
result, i.e. consider the VM network traffic need during VM placement?
You might want to keep an eye on the nested-resource-provider work
planned for Queens as it will give you better options to model your
resources:
https://blueprints.launchpad.net/nova/+spec/nested-resource-providers
Cheers,
gibi
BR,
Philippe
[1]
https://blueprints.launchpad.net/nova/+spec/bandwidth-as-scheduler-metric
[2] https://wiki.openstack.org/wiki/NetworkBandwidthEntitlement
[3]
https://openstack.nimeyo.com/80515/openstack-scheduling-bandwidth-resources-nic_bw_kb-resource
[4] https://docs.openstack.org/nova/latest/user/placement.html
[5]
http://specs.openstack.org/openstack/nova-specs/priorities/pike-priorities.html#placement
[6] https://review.openstack.org/#/c/473627/
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe:
openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
