About: Simple project to run deep learning inferencing on AWS Lambda to distinguish donuts from bagels and vadas (South Indian savory dish).
Try this app live at https://bit.ly/donutornot
The tech stack of this project is shown below:
The core of this project is within the donut_or_not/app folder. It contains the main.py Python file in which the FastAPI server is created and REST endpoints are defined. The models/export.pkl contains the deep learning model trained for this job. The templates subfolder contains the HTML files using Jinja2 templating syntax. We use Jinja to dynamically construct pieces of the web page without having to rely too much on Javascript.
In addition, the app folder has the Dockerfile needed to build this into a Docker image. We will use that image to first test the project locally, then push it up to AWS ECR. Finally, we will build the lambda function using this image on the ECR. The requirements3.txt has the dependencies specified. These are installed when we build the Docker image.
The donut_env folder is the virtualenv created for local development. This is precursor to determining what deps are needed to build the Docker image. Since this folder is in the gitignore, is it not part of this repo, but will be created if you follow the set up instructions given below.
The donut_or_not/events, donut_or_not/samconfig.yml, donut_or_not/template.yml are created by SAM as part of the init command. We customize the template.yml and specify all the endpoints that need to be exposed via AWS API Gateway.
The training.ipynb has the notebook used to train the DL model using FastAI. I used the 2019 course, lessons 1,2 for this.
The set up for this project consists of two parts. The first part here talks about the bits needed to push this model and inferencing function (built using FastAPI lib) to the cloud using AWS Lambdas. The second part talks about the bits needed to train the PyTorch - FastAI model. These two environments are different from each other. While they can be on the same machine (a GPU powered one), they don't have to necessarily be.
The inferencing would run on a Docker Image on the AWS Lambda FaaS platform. AWS recently (March 2020) announced support for Docker images for Lambdas. The documentation around this is slim. So we will use the AWS SAM CLI to give us a framework which we will flush out with the necessary code.
The AWS SAM CLI is used here to initialize a Python Docker Image based function. The CLI will create the necessary folders with app.py, a Dockerfile, a template.yml file for cloud formation (which is the mechanism to build this deployment stack on the cloud) among other misc files. It also gives tooling to invoke the function locally to allow the development cycle. Finally, it allows us to deploy the stack on the cloud by auto-generating the necessary S3 buckets (for storage), AWS ECR (Elastic Container Registry) entries to store the Docker image, the API gateway endpoints (for the client facing REST API) and gives us a public URL to invoke the function.
Thus, the prerequisites for this set up are
- AWS account - free tier. For this project, just the free tier will do. The first account you create is the "Root administrator" account. I recommend creating another account (based on security best practices), perhaps with admin privileges, but enable programming or shell access. We will use this account for this project.
- AWS CLI. After installation, you need to login to allow programmatic access. See this help.
- Docker desktop engine to build your containers locally.
- AWS SAM CLI. This is different from the AWS CLI. The SAM CLI makes use of the AWS CLI internally.
Use terminal to enter your project folder and then run sam init. SAM uses a guided approach to ask a series of questions and creates the necessary files as shown below. Note: If you are cloning this repo, you might get these files pre-made.
(base) ➜ donut_or_not git:(main) sam init
Which template source would you like to use?
1 - AWS Quick Start Templates
2 - Custom Template Location
Choice: 1
What package type would you like to use?
1 - Zip (artifact is a zip uploaded to S3)
2 - Image (artifact is an image uploaded to an ECR image repository)
Package type: 2
Which base image would you like to use?
1 - amazon/nodejs14.x-base
2 - amazon/nodejs12.x-base
3 - amazon/nodejs10.x-base
4 - amazon/python3.9-base
5 - amazon/python3.8-base
6 - amazon/python3.7-base
7 - amazon/python3.6-base
8 - amazon/python2.7-base
9 - amazon/ruby2.7-base
10 - amazon/ruby2.5-base
11 - amazon/go1.x-base
12 - amazon/java11-base
13 - amazon/java8.al2-base
14 - amazon/java8-base
15 - amazon/dotnet5.0-base
16 - amazon/dotnetcore3.1-base
17 - amazon/dotnetcore2.1-base
Base image: 5
Project name [sam-app]: donut_or_not
Cloning from https://github.com/aws/aws-sam-cli-app-templates
AWS quick start application templates:
1 - Hello World Lambda Image Example
2 - PyTorch Machine Learning Inference API
3 - Scikit-learn Machine Learning Inference API
4 - Tensorflow Machine Learning Inference API
5 - XGBoost Machine Learning Inference API
Template selection: 2
-----------------------
Generating application:
-----------------------
Name: donut_or_not
Base Image: amazon/python3.8-base
Dependency Manager: pip
Output Directory: .
Next steps can be found in the README file at ./donut_or_not/README.md
SAM CLI update available (1.33.0); (1.31.0 installed)
To download: https://docs.aws.amazon.com/serverless-application-model/latest/developerguide/serverless-sam-cli-install.htmlThis creates another folder by name donut_or_not in the current folder. There is an app, events folders, a template.yml and a training.ipynb file. For now, these files correspond to a simple MNIST example using PyTorch. We will delete or replace these files with ones needed for our workflow.
For the dev cycle, we need a faster build-deploy loop. For this, we will create a local virtual env.
(base) ➜ donut_or_not git:(main) ✗ conda activate aws_lambda_default
(aws_lambda_default) ➜ donut_or_not git:(main) ✗ python -m venv donut_env
(aws_lambda_default) ➜ donut_or_not git:(main) ✗ . donut_env/bin/activate
# Start installing libs for the web server
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install fastapi
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install uvicorn[standard]
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install jinja2
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install mangum
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install aiofiles
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install python-multipart
# Install libs for DL inference
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip install fastai==1.0.61
# Write the requirements.txt file
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ pip freeze > ./donut_or_not/app/requirements.txtSee donut_or_not_training.ipynb notebook to understand how the model was trained. Toward the end, you can see how the model was exported to a pickle while, which we use here.
The classify_img function in main.py file shows how the learner is hydrated from the model and how the uploaded image is used for inference.
The main.py is written in such a way that it can be run locally (without even the Docker container env) using the donut_env using the same code. It would work just as well when run in the Docker container later when testing using SAM. To run it locally, from within the app folder, run
(donut_env) (aws_lambda_default) ➜ app git:(main) ✗ uvicorn main:app --reload
INFO: Will watch for changes in these directories: ['/.../donut_or_not/donut_or_not/app']
INFO: Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
INFO: Started reloader process [7138] using watchgod
INFO: Started server process [7140]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: 127.0.0.1:54397 - "GET / HTTP/1.1" 200 OK
Once the app looks good, we can get on to the next stage, which is building the Docker image.
Now that the local env works, the next step is to build the Docker image. We use sam build command for this.
The inference environments for DL apps like this one are finicky and hard to get right if we were to just go with requirements or environment files. Hence, the developer needs to take the onus of building the Docker image and deploying it on the server. The build instructions are maintained in the app/Dockerfile file. After much trail and error, I settled with the current mix of install steps split between the Dockerfile and ones specified in the app/requirements3.txt file. To build, you run
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ sam build
Building codeuri: /Users/atma6951/Documents/code/pychakras/donut_or_not/donut_or_not runtime: None metadata: {'Dockerfile': 'Dockerfile', 'DockerContext': '/Users/atma6951/Documents/code/pychakras/donut_or_not/donut_or_not/app', 'DockerTag': 'python3.8-v1'} functions: ['DonutOrNotFunction']
Building image for DonutOrNotFunction function
Setting DockerBuildArgs: {} for DonutOrNotFunction function
Step 1/8 : FROM public.ecr.aws/lambda/python:3.8
---> 80342c69b467
Step 2/8 : COPY main.py ${LAMBDA_TASK_ROOT}
---> Using cache
---> 3ddf00ad5799
Step 3/8 : COPY requirements3.txt ${LAMBDA_TASK_ROOT}
---> 53661dd9063b
Step 4/8 : COPY templates ${LAMBDA_RUNTIME_DIR}/templates
---> b820476212a5
Step 5/8 : COPY models ${LAMBDA_RUNTIME_DIR}/models
---> cd972485a6af
Step 6/8 : RUN python3.8 -m pip install -r requirements3.txt -t "${LAMBDA_TASK_ROOT}"
---> Running in fa12b066d68c
Looking in links: https://download.pytorch.org/whl/torch_stable.html
Collecting torch==1.8.0+cpu
Downloading https://download.pytorch.org/whl/cpu/torch-1.8.0%2Bcpu-cp38-cp38-linux_x86_64.whl (169.1 MB)
Collecting torchvision==0.9.0+cpu
Downloading https://download.pytorch.org/whl/cpu/torchvision-0.9.0%2Bcpu-cp38-cp38-linux_x86_64.whl (13.3 MB)
Collecting aiofiles==0.7.0
Downloading aiofiles-0.7.0-py3-none-any.whl (13 kB)
# Install truncated for brevity here
Installing collected packages: typing-extensions, six, numpy, urllib3, torch, tomli, starlette, soupsieve, regex, pytz, python-dateutil, pyparsing, pydantic, platformdirs, Pillow, pathspec, mypy-extensions, MarkupSafe, kiwisolver, idna, h11, cycler, click, charset-normalizer, certifi, asgiref, websockets, watchgod, uvloop, uvicorn, torchvision, scipy, requests, PyYAML, python-multipart, python-dotenv, pynvx, pandas, packaging, nvidia-ml-py3, numexpr, matplotlib, mangum, Jinja2, httptools, fastprogress, fastapi, black, beautifulsoup4, aiofiles
Running setup.py install for python-multipart: started
Running setup.py install for python-multipart: finished with status 'done'
Running setup.py install for nvidia-ml-py3: started
Running setup.py install for nvidia-ml-py3: finished with status 'done'
Successfully installed Jinja2-3.0.2 MarkupSafe-2.0.1 Pillow-8.3.2 PyYAML-5.4.1 aiofiles-0.7.0 asgiref-3.4.1 beautifulsoup4-4.10.0 black-21.9b0 certifi-2021.5.30 charset-normalizer-2.0.6 click-8.0.1 cycler-0.10.0 fastapi-0.68.1 fastprogress-1.0.0 h11-0.12.0 httptools-0.2.0 idna-3.2 kiwisolver-1.3.2 mangum-0.12.2 matplotlib-3.4.3 mypy-extensions-0.4.3 numexpr-2.7.3 numpy-1.21.2 nvidia-ml-py3-7.352.0 packaging-21.0 pandas-1.3.3 pathspec-0.9.0 platformdirs-2.4.0 pydantic-1.8.2 pynvx-1.0.0 pyparsing-2.4.7 python-dateutil-2.8.2 python-dotenv-0.19.0 python-multipart-0.0.5 pytz-2021.3 regex-2021.9.30 requests-2.26.0 scipy-1.7.1 six-1.16.0 soupsieve-2.2.1 starlette-0.14.2 tomli-1.2.1 torch-1.8.0+cpu torchvision-0.9.0+cpu typing-extensions-3.10.0.2 urllib3-1.26.7 uvicorn-0.15.0 uvloop-0.16.0 watchgod-0.7 websockets-10.0
WARNING: Running pip as root will break packages and permissions. You should install packages reliably by using venv: https://pip.pypa.io/warnings/venv
WARNING: You are using pip version 21.1.1; however, version 21.2.4 is available.
You should consider upgrading via the '/var/lang/bin/python3.8 -m pip install --upgrade pip' command.
---> e13ae5fb8fbf
Step 7/8 : RUN python3.8 -m pip install --no-deps fastai==1.0.61 -t "${LAMBDA_TASK_ROOT}"
---> Running in d1f6b0692068
Collecting fastai==1.0.61
Downloading fastai-1.0.61-py3-none-any.whl (239 kB)
Installing collected packages: fastai
Successfully installed fastai-1.0.61
WARNING: Running pip as root will break packages and permissions. You should install packages reliably by using venv: https://pip.pypa.io/warnings/venv
WARNING: You are using pip version 21.1.1; however, version 21.2.4 is available.
You should consider upgrading via the '/var/lang/bin/python3.8 -m pip install --upgrade pip' command.
---> fa382927593b
Step 8/8 : CMD ["main.handler"]
---> Running in 5d3ee69f73a9
---> 0dd800d5144d
Successfully built 0dd800d5144d
Successfully tagged donutornotfunction:python3.8-v1
Build Succeeded
Built Artifacts : .aws-sam/build
Built Template : .aws-sam/build/template.yaml
Commands you can use next
=========================
[*] Invoke Function: sam local invoke
[*] Deploy: sam deploy --guided
You can test the image locally by running sam local start-api as shown below:
(donut_env) (aws_lambda_default) ➜ donut_or_not git:(main) ✗ sam local start-api
Mounting DonutOrNotFunction at http://127.0.0.1:3000/hello [GET]
Mounting DonutOrNotFunction at http://127.0.0.1:3000/ [GET]
Mounting DonutOrNotFunction at http://127.0.0.1:3000/getTimestamp [GET]
Mounting DonutOrNotFunction at http://127.0.0.1:3000/listImgFiles [GET]
Mounting DonutOrNotFunction at http://127.0.0.1:3000/classifyImg [POST]
You can now browse to the above endpoints to invoke your functions. You do not need to restart/reload SAM CLI while working on your functions, changes will be reflected instantly/automatically. You only need to restart SAM CLI if you update your AWS SAM template
2021-10-08 14:03:28 * Running on http://127.0.0.1:3000/ (Press CTRL+C to quit)
Invoking Container created from donutornotfunction:python3.8-v1
Building image.................
Skip pulling image and use local one: donutornotfunction:rapid-1.31.0.
START RequestId: 774ffad2-5327-41de-8d01-7977c56e4016 Version: $LATEST
END RequestId: 774ffad2-5327-41de-8d01-7977c56e4016
REPORT RequestId: 774ffad2-5327-41de-8d01-7977c56e4016 Init Duration: 0.11 ms Duration: 491.71 ms Billed Duration: 500 ms Memory Size: 128 MB Max Memory Used: 128 MB
2021-10-08 14:03:34 127.0.0.1 - - [08/Oct/2021 14:03:34] "GET / HTTP/1.1" 200 -
Invoking Container created from donutornotfunction:python3.8-v1
Building image.................
Skip pulling image and use local one: donutornotfunction:rapid-1.31.0.
To deploy you run sam deploy. The first time, you can use a guided approach and run sam deploy --guided. In my case, I have multiple AWS IAM users. Hence I provide a profile that has the right privileges. During the deploy, SAM walks you through various questions. Most are safe to answer with a y or Y for yes. The process creates necessary S3 buckets, repositories in ECR, IAM users and appropriate API gateways are opened.
(base) ➜ donut_or_not git:(main) sam deploy --guided --profile atma_lambda
Configuring SAM deploy
======================
Looking for config file [samconfig.toml] : Not found
Setting default arguments for 'sam deploy'
=========================================
Stack Name [sam-app]: donut-or-not
AWS Region [us-west-2]:
#Shows you resources changes to be deployed and require a 'Y' to initiate deploy
Confirm changes before deploy [y/N]: Y
#SAM needs permission to be able to create roles to connect to the resources in your template
Allow SAM CLI IAM role creation [Y/n]: Y
DonutOrNotFunction may not have authorization defined, Is this okay? [y/N]: Y
DonutOrNotFunction may not have authorization defined, Is this okay? [y/N]: Y
DonutOrNotFunction may not have authorization defined, Is this okay? [y/N]: Y
DonutOrNotFunction may not have authorization defined, Is this okay? [y/N]: Y
DonutOrNotFunction may not have authorization defined, Is this okay? [y/N]: Y
Save arguments to configuration file [Y/n]: Y
SAM configuration file [samconfig.toml]:
SAM configuration environment [default]:
Looking for resources needed for deployment:
Managed S3 bucket: aws-sam-cli-managed-default-samclisourcebucket-e7w5utol4v72
A different default S3 bucket can be set in samconfig.toml
Image repositories: Not found.
#Managed repositories will be deleted when their functions are removed from the template and deployed
Create managed ECR repositories for all functions? [Y/n]: Y
Saved arguments to config file
Running 'sam deploy' for future deployments will use the parameters saved above.
The above parameters can be changed by modifying samconfig.toml
Learn more about samconfig.toml syntax at
https://docs.aws.amazon.com/serverless-application-model/latest/developerguide/serverless-sam-cli-config.html
The push refers to repository [939098422637.dkr.ecr.us-west-2.amazonaws.com/donutornot3159b213/donutornotfunction89144d83repo]
The push refers to repository [939098422637.dkr.ecr.us-west-2.amazonaws.com/donutornot3159b213/donutornotfunction89144d83repo]
e8463ca45505: Pushed
c29f53611dcb: Pushed
b1cb94c6d9d1: Pushed
8030ec00e438: Pushed
e6ef369ea50f: Pushed
9b828246826b: Pushed
a1f8e0568112: Pushed
bcf453d1de13: Pushed
f6ae2f36d5d7: Pushed
5959c8f9752b: Pushed
3e5452c20c48: Pushed
9c4b6b04eac3: Pushed
donutornotfunction-0dd800d5144d-python3.8-v1: digest: sha256:f6df036750adbe9a39e4486383fb3ea094558f3e210182523f3c2f276840d77d size: 2841
Deploying with following values
===============================
Stack name : donut-or-not
Region : us-west-2
Confirm changeset : True
Deployment image repository :
{
"DonutOrNotFunction": "939098422637.dkr.ecr.us-west-2.amazonaws.com/donutornot3159b213/donutornotfunction89144d83repo"
}
Deployment s3 bucket : aws-sam-cli-managed-default-samclisourcebucket-e7w5utol4v72
Capabilities : ["CAPABILITY_IAM"]
Parameter overrides : {}
Signing Profiles : {}
Initiating deployment
=====================
Uploading to donut-or-not/4eb668d1f8cd57d339f709d83932edda.template 1635 / 1635 (100.00%)
Waiting for changeset to be created..
CloudFormation stack changeset
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Operation LogicalResourceId ResourceType Replacement
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+ Add DonutOrNotFunctionClassifyImgPermissionProd AWS::Lambda::Permission N/A
+ Add DonutOrNotFunctionGenTimeStampPermissionProd AWS::Lambda::Permission N/A
+ Add DonutOrNotFunctionHelloPermissionProd AWS::Lambda::Permission N/A
+ Add DonutOrNotFunctionListImgFilesPermissionProd AWS::Lambda::Permission N/A
+ Add DonutOrNotFunctionRole AWS::IAM::Role N/A
+ Add DonutOrNotFunctionRootPermissionProd AWS::Lambda::Permission N/A
+ Add DonutOrNotFunction AWS::Lambda::Function N/A
+ Add ServerlessRestApiDeployment53b600a676 AWS::ApiGateway::Deployment N/A
+ Add ServerlessRestApiProdStage AWS::ApiGateway::Stage N/A
+ Add ServerlessRestApi AWS::ApiGateway::RestApi N/A
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Changeset created successfully. arn:aws:cloudformation:us-west-2:939098422637:changeSet/samcli-deploy1633729019/6fc074ea-cd21-4ad6-bd01-195f7a631170
Previewing CloudFormation changeset before deployment
======================================================
Deploy this changeset? [y/N]: y
2021-10-08 15:05:38 - Waiting for stack create/update to complete
CloudFormation events from changeset
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
ResourceStatus ResourceType LogicalResourceId ResourceStatusReason
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
CREATE_IN_PROGRESS AWS::IAM::Role DonutOrNotFunctionRole Resource creation Initiated
CREATE_IN_PROGRESS AWS::IAM::Role DonutOrNotFunctionRole -
CREATE_COMPLETE AWS::IAM::Role DonutOrNotFunctionRole -
CREATE_IN_PROGRESS AWS::Lambda::Function DonutOrNotFunction -
CREATE_IN_PROGRESS AWS::Lambda::Function DonutOrNotFunction Resource creation Initiated
CREATE_COMPLETE AWS::Lambda::Function DonutOrNotFunction -
CREATE_IN_PROGRESS AWS::ApiGateway::RestApi ServerlessRestApi -
CREATE_IN_PROGRESS AWS::ApiGateway::RestApi ServerlessRestApi Resource creation Initiated
CREATE_COMPLETE AWS::ApiGateway::RestApi ServerlessRestApi -
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionClassifyImgPermissionProd -
CREATE_IN_PROGRESS AWS::ApiGateway::Deployment ServerlessRestApiDeployment53b600a676 -
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionGenTimeStampPermissionProd -
CREATE_IN_PROGRESS AWS::ApiGateway::Deployment ServerlessRestApiDeployment53b600a676 Resource creation Initiated
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionListImgFilesPermissionProd -
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionRootPermissionProd Resource creation Initiated
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionHelloPermissionProd Resource creation Initiated
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionClassifyImgPermissionProd Resource creation Initiated
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionHelloPermissionProd -
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionRootPermissionProd -
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionGenTimeStampPermissionProd Resource creation Initiated
CREATE_COMPLETE AWS::ApiGateway::Deployment ServerlessRestApiDeployment53b600a676 -
CREATE_IN_PROGRESS AWS::Lambda::Permission DonutOrNotFunctionListImgFilesPermissionProd Resource creation Initiated
CREATE_IN_PROGRESS AWS::ApiGateway::Stage ServerlessRestApiProdStage -
CREATE_IN_PROGRESS AWS::ApiGateway::Stage ServerlessRestApiProdStage Resource creation Initiated
CREATE_COMPLETE AWS::ApiGateway::Stage ServerlessRestApiProdStage -
CREATE_COMPLETE AWS::Lambda::Permission DonutOrNotFunctionGenTimeStampPermissionProd -
CREATE_COMPLETE AWS::Lambda::Permission DonutOrNotFunctionListImgFilesPermissionProd -
CREATE_COMPLETE AWS::Lambda::Permission DonutOrNotFunctionClassifyImgPermissionProd -
CREATE_COMPLETE AWS::Lambda::Permission DonutOrNotFunctionRootPermissionProd -
CREATE_COMPLETE AWS::Lambda::Permission DonutOrNotFunctionHelloPermissionProd -
CREATE_COMPLETE AWS::CloudFormation::Stack donut-or-not -
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
CloudFormation outputs from deployed stack
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Outputs
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key DonutOrNotFunctionApi
Description API Gateway endpoint URL for Prod stage for DonutOrNot function
Value https://r4eajdv9i1.execute-api.us-west-2.amazonaws.com/Prod/
Key DonutOrNotFunctionIamRole
Description Implicit IAM Role created for DonutOrNot Function
Value arn:aws:iam::939098422637:role/donut-or-not-DonutOrNotFunctionRole-11RF2YZ088JE4
Key DonutOrNotFunction
Description DonutOrNot ARN
Value arn:aws:lambda:us-west-2:939098422637:function:donut-or-not-DonutOrNotFunction-lomdtdVWL2UY
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Successfully created/updated stack - donut-or-not in us-west-2
The Lambda is not published at https://r4eajdv9i1.execute-api.us-west-2.amazonaws.com/Prod/ and can be accessed via any browser (mobile or desktop).
There are a few gotchas to keep in mind when running a web server via Lambda
- The API gateway configured on the cloud vs local runtime is quite different. If you want your API to accept uploaded files, you need to add
BinaryMediaTypes: ['*~1*']
to the template.yaml file. This allows the app to accept binary files of all types that are uploaded by the user. Without this, any upload command will be blocked, but return an un-helpful Internal Server Error response.
- Lambda runs your container with a different set of permission on the cloud compared to the local runtime. This can lead to all sorts of permission denial issues. You can address this by elevating permissions for certain folders or files using the
chmodcommand. See theDockerfilefor examples.