Deployment

A CloudFormation template deploys a complete Neptune Analytics + SageMaker notebook environment with a single command.

What it creates

Neptune Analytics graph with configurable provisioned memory
SageMaker notebook instance with nx_neptune pre-installed
S3 staging bucket with versioning + KMS encryption
KMS key for S3 bucket encryption
IAM role and policy with Neptune, S3, KMS, Athena, Glue, and SageMaker permissions

Notebook environment variables

Automatically set on the notebook instance:

Variable	Value
`NETWORKX_GRAPH_ID`	Created graph’s ID
`AWS_REGION`	Stack region
`NETWORKX_S3_EXPORT_BUCKET_PATH`	`s3://<bucket>/export/`
`NETWORKX_S3_IMPORT_BUCKET_PATH`	`s3://<bucket>/import/`
`NETWORKX_STAGING_BUCKET`	`s3://<bucket>/staging`

Quick deploy

Default deployment (installs from PyPI):
Terminal window
```
./cloudformation-templates/deploy.sh
```

Custom stack name and region:

./cloudformation-templates/deploy.sh my-stack us-east-1

Deploy with local wheel:

./cloudformation-templates/deploy.sh nx-neptune-demo us-west-1 true

Script arguments

Argument	Position	Default	Description
`STACK_NAME`	1st	`nx-neptune-demo`	Stack name (max 16 chars)
`REGION`	2nd	`us-west-1`	AWS region
`BUILD_WHEEL`	3rd	`false`	Build and deploy local wheel

Manual deploy with AWS CLI

# Build wheel and zip notebooks
python -m pip wheel -w dist .
zip -r /tmp/notebooks.zip notebooks/

# Upload to S3
aws s3 cp dist/nx_neptune-*.whl s3://your-bucket/nx-neptune/
aws s3 cp /tmp/notebooks.zip s3://your-bucket/nx-neptune/

# Deploy
aws cloudformation deploy \
  --stack-name nx-neptune-demo \
  --template-file cloudformation-templates/nx-neptune-sagemaker.json \
  --capabilities CAPABILITY_NAMED_IAM \
  --parameter-overrides AssetsS3Prefix=s3://your-bucket/nx-neptune

Parameters

Parameter	Description	Default
`ApplicationId`	Resource naming prefix (max 16 chars)	`nx-neptune`
`ProvisionedMemory`	m-NCUs for the graph (16, 32, 64)	`16`
`PublicConnectivity`	Enable public connectivity	`true`
`NotebookInstanceType`	SageMaker instance type	`ml.t3.medium`
`AssetsS3Prefix`	S3 prefix with `notebooks.zip` and optional `.whl`	(required)

Outputs

Output	Description
`GraphId`	Neptune Analytics graph ID
`NotebookURL`	SageMaker notebook URL
`StagingBucketName`	S3 bucket for import/export

aws cloudformation describe-stacks --stack-name nx-neptune-demo --query 'Stacks[0].Outputs'

Teardown

./cloudformation-templates/teardown.sh                      # defaults
./cloudformation-templates/teardown.sh my-stack us-east-1   # custom

Notes

Environment variables require a notebook stop/start to update
On every notebook start, OnStart re-installs nx_neptune (from the .whl if provided, otherwise from PyPI)