A client library for accessing Oblivious OpenAPIs
Project description
PyOblv
A Python library to access Oblivious OpenAPIs.
Usage
First, create a client, with your API key. This key can be created in the Oblivious Console UI. Go to the settings page, and create your ApiKey. Once done, the client can be created as follows
from oblv import authenticate
client = authenticate(apikey=*your_key_here*)
Using this client, all the Oblivious actions can be performed, including the deployment of enclaves. Below is a step-by-step guide, to creating your first deployment.
1 Create a service
A service refers to a branch/tag of your repository that needs to be deployed. For your first service. you can use the repository ObliviousAI/FastAPI-Enclave-Services. Execute the below command to create the service.
client.add_service(repo_owner="ObliviousAI",repo_name="FastAPI-Enclave-Services", ref="master")
{ "message": "Success", "service": { "ref": "master", "service_validated": true, "sha": "6b1b3e9cf6b0cb7264aad2fc80d91a009ccf0fc1", "type": "branch" } }
You have now successfully created your first service, and it can be now deployed to an enclave.
2 Enclave Creation
To create an enclave, you need to decide on a few of the parameters -
- deployment_name - A unique deployment name. You cannot have two running deployments with the same name.
- region_name - could be one of the AWS regions from
- us-east-1 (N. Virginia)
- us-west-2 (Oregon)
- eu-central-1 (Frankfurt)
- eu-west-2 (London)
- visibility - private or public
- is_dev_env - True or False.
- tags - A list of tags for your deployment
- account_type - This is the VCS type, and only GitHub*** is supported for now.
- build_args - Arguments specific to your deployment. It includes adding users to their roles with their public keys, runtime arguments and
any additional arguments for your build. It also includes the infra
size needed for the deployment. The options are -
- c5.xlarge - CPU:4 RAM:8 GB -- Credit Utilization: 68.0/hr
- m5.xlarge - CPU:4 RAM:16 GB -- Credit Utilization: 76.8/hr
- r5.xlarge - CPU:4 RAM:32 GB -- Credit Utilization: 100.8/hr
- c5.2xlarge - CPU:8 RAM:16 GB -- Credit Utilization: 136.0/hr
- m5.2xlarge - CPU:8 RAM:32 GB -- Credit Utilization: 153.6/hr
A valid example for build_args for the service you created -
args = {
"users": {
"admin": [
{
"user_name": "<your name>",
"public key": "<your public key>"
}
]
},
"infra_reqs": "m5.xlarge",
}
📝 Note:
admin here is the role defined for the deployment in service.yaml file
The service used, does not have requirements for any runtime or build arguments, so not needed here.
from oblv.models import CreateDeploymentInput
input = CreateDeploymentInput(
owner="ObliviousAI",
repo="FastAPI-Enclave-Services",
account_type="github",
ref="master",
region_name="us-east-1",
deployment_name="Depl",
visibility="private",
is_dev_env=True,
tags=[],
build_args=args
)
response = client.create_deployment(input)
On successful request, enclave creation is initiated, and it returns an id for the deployment, which can be later used to track the status and connection details for the enclave.
To check the state of the deployment, run
client.deployment_info(response.deployment_id).current_state
'CFT Initiated'
When the deployment state becomes Running, it indicates that the enclave is now available for connection.
To connect to the enclave, you need Oblv CLI installed in your system. The commands to use the CLI can be found in this documentation.
Cli binaries can be found here
The URL to connect to the enclave using CLI can be accessed using
client.deployment_info(response.deployment_id).instance.service_url
'https://conso-appli-15aiaxip9pcxk-94364694.enclave.oblivious.ai'
The PCR codes can be found using
client.deployment_info(response.deployment_id).pcr_codes
['d8dd856bacf4a8f0840ab530579b906091803e818e01dc4b8f51c247a7adfcfe55b4ef5f17be6d53bb952d92d87a376c', 'bcdf05fefccaa8e55bf2c8d6dee9e79bbff31e34bf28a99aa19e6b29c37ee80b214a414b7607236edf26fcb78654e63f','79477bad9504a347afb557a8ccd6f62ea61243311dff39f66944f5150242128da0cd070f0ce629c6beb6bb4f0b52f5ed' ]
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.