[Engine-devel] Feedback on ovirt-engine-sdk

[Michael Pasternak]

Hi Geert,

On 11/26/2011 07:49 PM, Geert Jansen wrote:
> Hi Michael,
> 
> i just wrote a small program with the new Python SDK. I found it very easy to use, so great work. I do have a few points of feedback:
> 
>  * Probably we want all Exceptions derived from a single base class. That way, the user can intercept any ovirtsdk with a single try/except clause. Currently that's not
> possible, as all your exceptions derive directly from Exception.

sure, in my TODO list.

> 
>  * I think the .get() method should be id-based not name based. Using "id" as a primary key has a number of advantages over name:
> 
>   - It is immutable, which makes e.g. renaming a lot easier.
>   - It allows you to retrieve an object without a search query or a filter.
>   - It matches how the API works as well.

"id" is supported, you can do api.collection.get(id=xxx), only as you mentioned
"id" is not primary, but secondary key (i thought that searching by name will be easier),
but i can change it in a second if this is an issue.

> 
>   Also currently i think it's not possible to retrieve a resource by its ID, unless you use the id= keyword argument to list() which retrieves the whole collection and then
> filters out the matching id.

nope, it's supported - i even added it to examples.py [1]
-> vm5 = api.vms.get(id='02f0f4a4-9738-4731-83c4-293f3f734782')

[1] will add it to get method .get(id, name) as you suggested.

> 
>  * You may want to introduce some non-entity related operations, like ping(), and connect(). These allow you to better control when you connect, and allow an program to
> test if the connection details work. That helps client except the right exceptions at the right time, and display more relevant error messages.

- no need for explicit connect() as its done implicitly on proxy construction
- no need for ping() as any attempt to use disconnected (and not recoverable) proxy, will cause ConnectionError,
  so user should only /Trap/ this exception for disaster_recovery

> 
>  * There's two Python objects involved for each API object. For a VM for example, you have params.VM and brokers.VM. This distinction is made visible to the user, for example:
> 
>   template = api.templates.get('Blank')
> 
>   vm = params.VM()
>   vm.name = 'foo'
>   vm.template = template  # ERROR

yep, you right doing this is in my TODO list as well
(had to switch to CLI, so i do have few leftovers[1] - this is one of them)

[1] they all have simple workarounds (like this one as described below)

> 
> The last line need to be:
> 
>   vm.template = params.Template(id=template.id)

for now (as workaround you can pass 'superclass', e.g vm.template = template.superclass),
but i suggest you just wait for another day till I'll support this.

> 
> In my version of the Python API i used mix-in classes to add the OO behavior to the generated classes. It was easy, because PyXB had support for that. This gave me the
> single class behavior. I am not sure if GenerateDS generated code allows that. I'm pretty sure it could be done though, but you may need to do some Python magic.

no need, i do this with regular inheritance pattern

> 
>  * Nitpick: should this just be called "ovirt" instead of "ovirtsdk"?

initially it was, but since both cli & sdk would operate in same namespace,
i wanted to avoid future collisions and renamed them to ovirtsdk/ovirtcli.

> 
> Overall great work though!

10nx a lot mate.

> 
> Regards,
> Geert
> _______________________________________________
> Engine-devel mailing list
> Engine-devel at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/engine-devel


-- 

Michael Pasternak
RedHat, ENG-Virtualization R&D