On 18/06/13 23:32 +0000, Adrian Otto wrote:
Yes. I think having a POST method in the API makes perfect sense. Assuming we
reach agreement on that, the next question that comes up is:
Err, I am not convinced.
Before that, I think it's worth highlighting the different
proposals/requirements here:
1) heat needs a way to setup an autoscaling group
2) autoscaling needs a way of telling heat to scale up/down
3) People might want to integrate other orchestration engines with Heat
So one by one:
1) autoscaling has a post api that the Heat autoscaling group resource
posts to, Heat will provide a webhook in the event of a scaling action.
2) when autoscaling determines that there should be a scaling action
it calls the webhook. The reason I suggest a webhook is in-instance
applications might want to scale their applications (like
openshift-gears) - so don't assume a heat endpoint.
Also we can have this as an action not a resource-create.
PUT /$tenant/stacks/$stack/resources/<autoscale-group>
(with an action of scale-up/down)
I think we should discourage (make it impossible) for users from modifing
stacks outside of stack-update.
To me one of the most powerful and apealing things of Heat is the
ability to reproducibly re-create a stack from a template. This
new public API is going to make this difficult.
Also creating a resource is not a trivial issue, the user would have
create the resources in the correct order (with correct inter-resource
references etc..) mostly throwing away the point of orchestation in
the first place. If you are doing this you may as well talk directly
to nova/cinder/networking directly.
How to do you modify resources that have been created with a POST?
You mention HTTP PUT as an answer to that. Unfortunately PUT is only really
useful for doing a full resource replacement, not just tweaking something
that's already there. For that, you really want HTTP PATCH
(http://tools.ietf.org/html/rfc5789). You can make this really elegant for JSON
with JSON Patch (http://tools.ietf.org/html/draft-ietf-appsawg-json-patch-10).
We should note that offering API methods to adjust a stack (aka: assembly)
means that there will be a divergence between what's described in the original
template, and the actual running state of the stack/assembly created by the
template, well beyond the results of an autoscale policy. In fact, it would be
possible to build a stack/assembly with no template at all, if the right API
methods are present. There are good use cases for this, particularly for higher
level compatibility layers where it would be awkward to generate permutations
of templates to immediately feed into an API, rather than just use an API
method for adjusting the stack/assembly in place. it would be much more
elegant, for example, to implement a CAMP implementation on top of Heat if Heat
had a REST API for creating and managing individual resources within a
stack/assembly. This same argument applies to integrating any other
orchestration or configuration management system with Heat.
So this is "3)":
As I was alluding to above, if you are integrating at this level what
value is Heat providing to you? If you are just using it for resource
create this is a thin layer over the python clients. I'd suggest just
using the resources as python plugins. I think exposing this would
bring more harm than good.
-Angus
More comments in-line:
On Jun 18, 2013, at 3:44 PM, Christopher Armstrong
<chris.armstr...@rackspace.com>
wrote:
tl;dr POST /$tenant/stacks/$stack/resources/ ?
Yes.
== background ==
While thinking about the Autoscaling API, Thomas Hervé and I had the
following consideration:
- autoscaling is implemented as a set of Heat Resources
- there are already general APIs for looking at resources generically:
- resource-show (GET /$tenant/stacks/$stack/resources/$id)
- resource-metadata (GET /$tenantt/stacks/$stack/resources/$id/metadata)
- resource-list (GET /$tenant/stacks/$stack/resources/)
- we want to be able to create and configure autoscaling resources
through the API
- maybe we should implement POST for resources?
This is basically the gist of the question. I believe the answer
should be the same as the answer about any other type of resource we
might want to manipulate through the API -- it seems best that either
all resource types are manipulated through a generic resource
manipulation API, or they should all have their own specific ReST
collection.
Give them specific collections, so they can be easily specialized.
Actually, I could also imagine a situation where only generic
operations on common resource metadata are allowed via
/$tenant/stacks/$stack/resources/, and resource-specific manipulation
is done via resource-specific collections -- I don't know how ReSTy
that is, though.
I'll get to specifics. There are two ways I can imagine the autoscale
API looking. I'll avoid the word "resource" when referring to ReST
resources and just talk about "collections" and "paths", since
"resource" in this context also means Heat resources.
== resource-specific paths ==
One is basically just like Otter's: http://docs.autoscale.apiary.io/
This provides paths like /$tenant/groups/$id (for an autoscaling
group), /$tenant/groups/$id/policies (for a policy), etc. These
variously support GET for reading as well as POST and PUT for
manipulation.
We can use "/v1.0/{tenantId}/groups/{groupId}/policies" as an example
operation. We POST JSON describing a new scaling policy to create to a
new scaling policy.
The above approach is definitely my preference.
== generic paths ==
The alternative is to say that autoscaling groups, policies, etc are
all Just Heat Resources, and Heat resources already have a ReST
collection at /$tenant/resources/.
In this option, the alternative to POSTing to
/$tenant/groups/$id/policies would be to post directly to
/$tenant/resources/, with a body exactly like in the previous example,
but with two more JSON attributes:
- the type of the resource, in this case something like
"AWS::AutoScaling::ScalingPolicy"
- the group ID that the new policy should be associated with, since
it's not specified in the URL.
This is approach is less elegant than resource-specific paths. Bringing in
types is a whole other layer of complexity that can be avoided when using
resource-specific paths. You actually still have types, but they are implicit.
You still have to document them, but you don't have to individually name them.
One concern I have is about how well we can specify a strict schema of
inputs and outputs to the resources/ collection -- I'm particular
interested in JSON hyperschema. I'm not sure how it handles
heterogeneous collections like this.
In the JSON world, that's what documentation is for. Although having a JSON
Schema can help simplify client development, it is yet another body of code
that Heat developers will need to maintain. The question here comes down to how
many developers will work on building clients, and could simplifying their
development experience help boost adoption of Heat. My suggestion is (even if
we like the idea of a schema) to start lean without a schema, and add that in
if we agree it's something we want later. With some discipline, it's possible
to implement things in a way that lends itself to using a schema even if we
don't explicitly define, publish, and maintain one.
Maintaining the schema (and the documentation for that matter) should be
something that all OpenStack development teams feel comfortable with, because
basically anytime someone changes any of the downstream OpenStack API's,
there's a good chance it will need to change here too in order to expose
whatever feature they add. A tweak to a documentation page may be less
burdensome if you think of it from that perspective.
Adrian
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
