Open Stack

On 05/24/2017 10:07 AM, Clark Boylan wrote:
>
>
> On Wed, May 24, 2017, at 07:39 AM, Matt Riedemann wrote:
>> Rocky tipped me off to a request to document config drive which came up
>> at the Boston Forum, and I tracked that down to Clark's wishlist
>> etherpad [1] (L195) which states:
>>
>> "Document the config drive. The only way I have been able to figure out
>> how to make a config drive is by either reading nova's source code or by
>> reading cloud-init's source code."
>>
>> So naturally I have some questions, and I'm looking to flesh the idea /
>> request out a bit so we can start something in the in-tree nova devref.
>>
>> Question the first: is this existing document [2] helpful? At a high
>> level, that's more about 'how' rather than 'what', as in what's in the
>> config drive.
>
> This is helpful, but I think it is targeted to the deployer of OpenStack
> and not the consumer of OpenStack.

Yup.

>> Question the second: are people mostly looking for documentation on the
>> content of the config drive? I assume so, because without reading the
>> source code you wouldn't know, which is the terrible part.
>
> I'm (due to being a cloud user) mostly noticing the lack of information
> on why cloud users might use config drive and how to consume it.
> Documentation for the content of the config drive is a major piece of
> what is missing. What do the key value pairs mean and how can I use them
> to configure my nova instances to operate properly.
>
> But also general information like, config drive can be more reliable
> that metadata service as its directly attached to instance. Trade off is
> possibly no live migration for the instance (under what circumstances
> does live migration work and as a user is that discoverable?). What
> filesystems are valid and I need to handle in my instance images? Will
> the device id always be config-2? and so on. The user guide doc you
> linked does try to address some of this, but seems to do so from the
> perspective of the person deploying a cloud, "do this if you want to
> avoid dhcp in your cloud", "install these things on compute hosts".

Yah. Being able to point consumers at why they should care and what 
benefits it has is useful since it's a flag in the server api to turn on 
config-drive.

But also ++ to clarkb

>> Based on this, I can think of a few things we can do:
>>
>> 1. Start documenting the versions which come out of the metadata API
>> service, which regardless of whether or not you're using it, is used to
>> build the config drive. I'm thinking we could start with something like
>> the in-tree REST API version history [3]. This would basically be a
>> change log of each version, e.g. in 2016-06-30 you got device tags, in
>> 2017-02-22 you got vlan tags, etc.
>
> I like this as it should enable cloud users to implement tooling that
> knows what it needs that can error properly if it ends up on a cloud too
> old to contain the required information.

++

>> 2. Start documenting the contents similar to the response tables in the
>> compute API reference [4]. For example, network_data.json has an example
>> response in this spec [5]. So have an example response and a table with
>> an explanation of fields in the response, so describe
>> ethernet_mac_address and vif_id, their type, whether or not they are
>> optional or required, and in which version they were added to the
>> response, similar to how we document microversions in the compute REST
>> API reference.
>
> ++

+100!

>>
>> --
>>
>> Are there other thoughts here or things I'm missing? At this point I'm
>> just trying to gather requirements so we can get something started. I
>> don't have volunteers to work on this, but I'm thinking we can at least
>> start with some basics and then people can help flesh it out over time.
>
> I like this, starting small to produce something useful then going from
> there makes sense to me.

++

> Another idea I've had is making a tool that collected (or was fed)
> information that goes into config drives and produces the device to
> attach to a VM would be nice. Reason for this is while config drive is
> something grown out of nova/OpenStack you often want to boot images with
> Nova and other tools so making it easy for those other tools to work
> properly too would be nice. In the simple case I build images locally,
> then boot them with kvm to test that they work before pushing things
> into OpenStack and config drive makes that somewhat complicated. Ideally
> this would be the same code that nova uses to generate the config drives
> just with a command line front end.

That would also help with testing config-drive consuming tools potentially.

We have some fixtures in the glean source repo:

http://git.openstack.org/cgit/openstack-infra/glean/tree/glean/tests/fixtures

which we collected from clouds out in the wild so we could make sure 
were were doing the right things from them. I imagine it could be neat 
to be able to use a tool to generate various combos of config-drive 
content on the fly so they don't have to be hard-coded- but if it's not 
the actual code itself it wouldn't be as awesome.

>>
>> [1] https://etherpad.openstack.org/p/openstack-user-api-improvements
>> [2] https://docs.openstack.org/user-guide/cli-config-drive.html
>> [3]
>> https://docs.openstack.org/developer/nova/api_microversion_history.html
>> [4] https://developer.openstack.org/api-ref/compute/
>> [5]
>> https://specs.openstack.org/openstack/nova-specs/specs/liberty/implemented/metadata-service-network-info.html#rest-api-impact
>
> Thank you for bringing this up,

Yes - thank you! I think this can be super helpful!

Monty