Skip to content

Latest commit

 

History

History
279 lines (231 loc) · 9.18 KB

xopera.md

File metadata and controls

279 lines (231 loc) · 9.18 KB
Raw

Deploy your functions

We have now successfully made a deployable model, only thing we need now is to deploy it.

RADON comes with its own deployment tool called xOpera.

The first thing needed is to export the model from GMT into the IDE.

Tasks

  • open GMT (link in the right side of the IDE)
  • Locate your model in the service templates, click on the template.
  • Click export and export to ide
  • Go back to the IDE and check that a new root folder called radon-csars has been created.
  • Open the folder and check that you have a file called the same as your service template and with the .csars extension.

You are now ready to either use the CLI version of the deployment tool, or the SaaS offering.

CLI version

xOpera aims to be a lightweight orchestrator compliant with OASIS TOSCA model used in RADON.

The orchestrator puts the application into the runtime environment, enforcing the state described by application blueprint (CSAR file downloaded) onto AWS. The common operations are deployment and un-deploy and are executed on different target environments such as staging, development and production.

The CLI can be

Tasks

Note: In order for this to work, you need to have set your AWS credentials as environment variables. For more info look at the env variables needed, and how to create IAM credentials

  • Download the .csar file from IDE, by clicking on the file in the IDE, and choose download
  • install opera through pip, or use the docker image in this repository.
pip

Look at the installation guide on the GH webpage

mkdir ~/opera && cd ~/opera
python3 -m venv .venv && . .venv/bin/activate
pip install opera boto3

Run the following command in the folder where you have the csar file stored (here called todo.csar)

opera deploy -c todo.csar

After executing the opera, you should see something like this printed out.

[Worker_0]   Deploying AwsPlatform_0_0
[Worker_0]     Executing configure on AwsPlatform_0_0
[Worker_0]   Deployment of AwsPlatform_0_0 complete
[Worker_0]   Deploying AwsDynamoDBTable_0_0
[Worker_0]     Executing create on AwsDynamoDBTable_0_0
[Worker_0]   Deployment of AwsDynamoDBTable_0_0 complete
[Worker_0]   Deploying AwsLambdaFunction_1_0
[Worker_0]     Executing create on AwsLambdaFunction_1_0
[Worker_0]   Deployment of AwsLambdaFunction_1_0 complete
[Worker_0]   Deploying AwsLambdaFunction_2_0
[Worker_0]     Executing create on AwsLambdaFunction_2_0
[Worker_0]   Deployment of AwsLambdaFunction_2_0 complete
[Worker_0]   Deploying AwsLambdaFunction_0_0
[Worker_0]     Executing create on AwsLambdaFunction_0_0
[Worker_0]   Deployment of AwsLambdaFunction_0_0 complete
[Worker_0]   Deploying AwsLambdaFunction_3_0
[Worker_0]     Executing create on AwsLambdaFunction_3_0
[Worker_0]   Deployment of AwsLambdaFunction_3_0 complete
[Worker_0]   Deploying AwsLambdaFunction_4_0
[Worker_0]     Executing create on AwsLambdaFunction_4_0
[Worker_0]   Deployment of AwsLambdaFunction_4_0 complete
[Worker_0]   Deploying AwsApiGateway_0_0
[Worker_0]     Executing create on AwsApiGateway_0_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_0_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_1_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_2_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_3_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_4_0
[Worker_0]     Executing configure on AwsApiGateway_0_0
[Worker_0]   Deployment of AwsApiGateway_0_0 complete
docker

In the same folder where you have the csar file, create a file called env.list with your AWS credentials like this:

AWS_ACCESS_KEY_ID=KEY_ID_HERE
AWS_SECRET_ACCESS_KEY=ACCESS_KEY_HERE

Run the following command in the folder where you have the csar file stored (here called todo.csar)

docker run -it -v $PWD:/tmp/opera --env-file env.list praqma/operacli opera deploy -c todo.csar
[Worker_0]   Deploying AwsPlatform_0_0
[Worker_0]     Executing configure on AwsPlatform_0_0
[Worker_0]   Deployment of AwsPlatform_0_0 complete
[Worker_0]   Deploying AwsDynamoDBTable_0_0
[Worker_0]     Executing create on AwsDynamoDBTable_0_0
[Worker_0]   Deployment of AwsDynamoDBTable_0_0 complete
[Worker_0]   Deploying AwsLambdaFunction_1_0
[Worker_0]     Executing create on AwsLambdaFunction_1_0
[Worker_0]   Deployment of AwsLambdaFunction_1_0 complete
[Worker_0]   Deploying AwsLambdaFunction_2_0
[Worker_0]     Executing create on AwsLambdaFunction_2_0
[Worker_0]   Deployment of AwsLambdaFunction_2_0 complete
[Worker_0]   Deploying AwsLambdaFunction_0_0
[Worker_0]     Executing create on AwsLambdaFunction_0_0
[Worker_0]   Deployment of AwsLambdaFunction_0_0 complete
[Worker_0]   Deploying AwsLambdaFunction_3_0
[Worker_0]     Executing create on AwsLambdaFunction_3_0
[Worker_0]   Deployment of AwsLambdaFunction_3_0 complete
[Worker_0]   Deploying AwsLambdaFunction_4_0
[Worker_0]     Executing create on AwsLambdaFunction_4_0
[Worker_0]   Deployment of AwsLambdaFunction_4_0 complete
[Worker_0]   Deploying AwsApiGateway_0_0
[Worker_0]     Executing create on AwsApiGateway_0_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_0_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_1_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_2_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_3_0
[Worker_0]     Executing pre_configure_source on AwsApiGateway_0_0--AwsLambdaFunction_4_0
[Worker_0]     Executing configure on AwsApiGateway_0_0
[Worker_0]   Deployment of AwsApiGateway_0_0 complete

After successfully deployed the model, head on to test the deployment in the section below.

SaaS version

NOTE: this is under heavy development, so the guide might not work with the latest version.

Opera SaaS steps:

  • Access the UI : https://xopera-radon.xlab.si/ui/. You can login using your radon credentials
  • Create a workspace
  • Create credentials
    • Name: whatever you want
    • For AWS use the same path that is used on local: /root/.aws/credentials
    • Copy the content of the credentials file found on your local machine on /root/.aws/credentials. Example below...
[default]
aws_access_key_id = <secret>
aws_secret_access_key = <secret>
  • Assign credentials to the workspace
  • Invoke opera SaaS deploy from IDE
  • Choose in which workspace to deploy it
  • Create a project providing a name
  • Choose whether you want to deploy to Cloud provider or not
  • Provide the Service template name ( you need to download the csar and unzip it. The service template name you need to provide looks like this: “_definitions/steIgeneral__myexampleproject.tosca”
  • Provide the inputs file.
    • It can be an empty inputs.yaml (containing :{}) and it can be stored in the IDE workspace. You access it in the projects folder on the popup
  • Check deployment status in the UI.
  • After successful deployment send a POST request to the apilink/todos with the following payload to tests the functionality of the application: { "todo": "Have a another coffee :-)" }

Test out the serverless application

Now comes the time to test out the serverless application.

You can either do it with the AWS UI, or with postman(or curl if you dare).

Click on the tool of choice below.

Tasks

AWS UI
  • Log into AWS account
  • Go into Amazon API Gateway and select ServerlessToDoListAPI
  • Click on the /todos GET method, and test and see that the result you get back in the response body is an empty list; []
  • To populate with a todo, Click on the /todos POST method, and test
  • In the resquest body, paste in the payload: {"todo":"buy milk"} and click test
  • Get back to test the /todos GET method to see that the item has been persisted.
Postman
  • Log into AWS account
  • Go into Amazon API Gateway and select ServerlessToDoListAPI
  • To get the URL to test, click on Stages production and take the invoke URL:
  • Open Postman
  • Send a get message to <invokeURL>/todos and see that the result you get back in the response body is an empty list; []
  • To populate with a todo, send a POST method to <invokeURL>/posts with a resquest body like {"todo":"buy milk"}
  • Get back to test the /todos GET method to see that the item has been persisted.

Clean up

When you are finished playing around, you can use opera to undeploy. To undeploy use opera undeploy and watch the logs to make sure that all gets deleted.