Lambda - Mobius External Grader

https://console.aws.amazon.com/lambda/home?region=us-east-1#/functions/Mobius_edx_Grader?tab=graph ARN: arn:aws:lambda:us-east-1:114056409474:function:Mobius_edx_Grader

Answer file template

for the .mob file to be uploaded to s3, its description should have some extra parameters indicating how the answer files are to be executed and graded. see more here.

Installation

Clone the project.
npm install to install the required packages.
If you do not have Typescript installed globally, do so with npm install -g typescript. This is required to build the project for uploading to Lambda.

Code Structure

src/core and src/libs folders are the modelling libraries in Mobius Parametric Modeller project.
src/model folder is from the core shared models in Mobius Parametric Modeller project
src/grader.ts is the main file to grade the student submission, with two grading functions: gradeFile_URL and gradeFile
- gradeFile_URL takes in a json structure of {"file": file_url}
- gradeFile takes in a json structure of {"file": file_string}

Building the code and Deploying

Lambda requires the code to be in javascript. So the codes have to be built before deploying on to Lambda.

Deploying using python script

The python file upload_to_amazon.py can be used to upload the project to amazon lambda.

Setting up

Make a copy of the AMAZON_KEY.template.py file and change its name to AMAZON_KEY.py. Fill in the file with your aws_access_key_id and aws_secret_access_key as well as your local mobius project directory (e.g. C:\\Users\\UserName\\Documents\\mobius-parametric-modeller-dev).
in upload_to_amazon.py, change FUNC_NAME variable to MAIN_FUNCTION or DEV_FUNCTION to indicate which lambda function for the project to be uploaded to.

Use the script

The script can be run with python (python upload_to_amazon.py or run with vs code)

Deploying manually

If for any reason the python script does not work, the project can be compiled and uploaded to amazon manually as follow:

Update the code

The project relies on a number of code from mobius, namely src/core and src/libs which are copied from src/assets/core and src/assets/libs respectively from mobius project.

Build the code

In a command prompt in the project folder, simply do tsc. A dist folder will be created with the built javascript code.

Deploy the code

Put the content of the dist folder into a .zip file. The name of the .zip file does not matter. Make sure that the .zip file structure is as follow (an example zip file can be found in zip_example folder - dist.zip):
```
. dist_zip_file.zip
└── core
|   └── ...
└── libs
|   └── ...
└── model
|   └── ...
└── grader.js
```
Go to the lambda function Configuration page, Function code panel, in the Code entry type dropdown menu, select "Upload a .zip file". Upload the .zip file and press Save on the top right corner of the page

External npm packages and deploying node_modules

External packages can be used in lambda. Just install them as per normal: npm install <<package>>

For the deployed lambda function to get access to these external packages, the node_modules folder must also be uploaded to lambda panel. Since the project's node_modules folder is quite large, if we are to upload the node_modules along with the dist folder, AWS will disable us from seeing the source code due to the big size. The way we're uploading node_modules onto AWS is through Lambda layers. It does not interfere with the main code. Also, we can use the same node_modules layer for multiple lambda functions if necessary. The process of uploading is as follow:

In the AWS lambda left side panel, choose layers. We can create a new layer or update an existing layer (create version).

upload the node_modules .zip file. the file must be in this specific structure (an example zip file can be found in zip_example folder - nodejs.zip):

. node_module_zip_file.zip
└── nodejs
    └── node_modules
        └── node package1
        |   └── ...
        └── node package2
        |   └── ...
        └── node package3
        |   └── ...
        └── ...

after uploading the file, go to the lambda function. In the Configuration page, Designer panel, select Layers. Add the newly uploaded layer and remove the old one.

More on this can be read here: https://medium.com/@anjanava.biswas/nodejs-runtime-environment-with-aws-lambda-layers-f3914613e20e

AWS API Gateway

The communication between the EC2 python server and the lambda function is done through API Gateway (which is attached to the lambda function) A basic tutorial for API Gateway can be found here https://docs.aws.amazon.com/apigateway/latest/developerguide/api-gateway-create-api-as-simple-proxy-for-lambda.html

For our lambda function, the only input required in the http request body is "file", which is a string of the submission file url (edx upload the file to their own database and provide us with a url for the file). This is set in the request body's model "MobiusFile".

design-automation/mobius-external-grader