[ODSC 51354] Create new Aqua deployment #551
Conversation
…ODSC-51354/create_new_deployment
oracle-contributor-agreement bot added the OCA Verified 
| 📌 Cov diff with main: 📌 Overall coverage: |
VipulMascarenhas requested review from lu-ohai, mayoor, mingkang111, mrDzurb and qiuosier VipulMascarenhas changed the title | 📌 Cov diff with main: No success to gather report. 😿 📌 Overall coverage: No success to gather report. 😿 |
| 📌 Cov diff with main: 📌 Overall coverage: |
| 📌 Cov diff with main: No success to gather report. 😿 📌 Overall coverage: No success to gather report. 😿 |
| 📌 Cov diff with main: No success to gather report. 😿 📌 Overall coverage: No success to gather report. 😿 |
| 📌 Cov diff with main: 📌 Overall coverage: |
| 📌 Cov diff with main: 📌 Overall coverage: |
| @dataclass | ||
| class AquaDeployment: | ||
| """Represents an Aqua Model Deployment""" | ||
| |
There was a problem hiding this comment.
during the design discussion we decided to have that in the aqua_service_model tag (ocid#model_name). We can have a separate field, but could get redundant due to the tag. Tracking this in a separate page here.
| else None | ||
| ) | ||
| if aqua_service_model: | ||
| results.append( |
There was a problem hiding this comment.
We can have something like
AquaDeployment.from_oci_model(...) | | ||
| try: | ||
| model_deployment = self.client.get_model_deployment(**kwargs).data | ||
| except ServiceError as se: |
There was a problem hiding this comment.
we add everywhere these exception statements, probably would be better to move them into the decorator?
There was a problem hiding this comment.
yes, we'll revisit this and update the code once the error handler is finalized.
| f"Target deployment {model_deployment.id} is not Aqua deployment." | ||
| ) | ||
| | ||
| return AquaDeployment( |
There was a problem hiding this comment.
NIT: AquaDeployment.from_oci_model()?
There was a problem hiding this comment.
yes, makes sense. will add it in a separate PR.
| raise HTTPError(400, Errors.NO_INPUT_DATA) | ||
| | ||
| # required input parameters | ||
| compartment_id = input_data.get("compartment_id") |
There was a problem hiding this comment.
For parameters it will be automatically handled, right?
| | ||
| # required input parameters | ||
| compartment_id = input_data.get("compartment_id") | ||
| if not compartment_id: |
There was a problem hiding this comment.
I would suggest to add some MDRequest dataclass and put all of the logic with constructing the input attributes and validation there.
| aqua_service_model = f"xxx.xxx.xxx.xxx.xxx#aqua_service_model_value" | ||
| | ||
| self.finish( | ||
| AquaDeploymentApp().create( |
There was a problem hiding this comment.
if we have MDRequest dataclass, then the code could be
AquaDeploymentApp().create(**md_request.to_dict()) | 📌 Cov diff with main: 📌 Overall coverage: |
| 📌 Cov diff with main: 📌 Overall coverage: |
3 participants
Description
Jira ticket: ODSC-51354
The code changes in this PR an Aqua model deployment. Following changes are done:
deployment_handler.py. The handler accepts post request and there are a few required parameters that caller will have to provide (need comments on what can be optional).deployment.py, implemented create() functionality. It does not create a catalog entry right now (needs AquaModel.create implemented - either mock or otherwise), so assumes a few attributes will be available (image, model id, entrypoint) but this will be removed. (need comments on what deployment config needs to be hardcoded vs. expected from caller).ads.common.serializer.DataClassSerializablefor CLI commands.User can invoke create by calling:
It should return a json response:
or, user can use cli:
Next steps: