GenU Updates and Parameter Changes
This guide explains how to update GenU and change parameters after deploying with one-click deployment. For detailed parameters supported by GenU, please refer to the GenU Documentation.
The following steps will be performed:
- Check auto-generated parameters from one-click deployment in AWS Systems Manager Parameter Store - Set up development environment with Amazon SageMaker Studio Code Editor - Use CDK for updates and parameter changes
Check Auto-Generated Parameters from One-Click Deployment
In one-click deployment, the parameters used during GenU deployment are stored in Parameter Store in JSON format.
Open the Parameter Store console (※This is the Tokyo region (ap-northeast-1) screen) and check the parameters. If deployed to a region other than Tokyo, check in the deployment destination region (e.g., us-east-1).
The following parameters are stored with the Environment name specified during deployment (default is dev)
- /genu/dev.json - All parameters for dev environment stored in JSON format
Tip
If deployed with staging or prod, they become /genu/staging.json and /genu/prod.json respectively.
The parameter content is stored in JSON format as follows:
{
"modelRegion": "us-east-1",
"modelIds": [
"us.anthropic.claude-sonnet-4-20250514-v1:0",
"us.anthropic.claude-opus-4-20250514-v1:0",
"us.deepseek.r1-v1:0",
"us.anthropic.claude-3-7-sonnet-20250219-v1:0",
"us.anthropic.claude-3-5-haiku-20241022-v1:0",
"us.amazon.nova-premier-v1:0",
"us.amazon.nova-pro-v1:0",
"us.amazon.nova-lite-v1:0",
"us.amazon.nova-micro-v1:0"
],
"ragKnowledgeBaseEnabled": true,
"selfSignUpEnabled": false,
"allowedSignUpEmailDomains": [
"gmail.com"
],
"allowedIpV4AddressRanges": null,
"allowedIpV6AddressRanges": null
}
Here's an example of the management console screen (turn on Show decrypted value to see the values):
Using this JSON data, you can check and change parameters by applying them to the corresponding environment section in the following steps.
Set Up Development Environment with SageMaker Code Editor
To update the GenU environment, we use SageMaker Code Editor. Create it using CloudFormation from the following link.
Tip
If you intentionally changed the GenU deployment region from the default Tokyo, also change the SageMaker Code Editor deployment destination region. If you haven't changed it intentionally, please deploy to the Tokyo region from the URL below.
Warning
Regarding pricing, when running the default ml.t3.medium in the Tokyo region, $0.065 per hour is charged. Code Editor automatically stops when no operations are performed for a certain period, providing cost optimization.
CloudFormation will open, so check the box at the bottom of the screen and press Create stack.
Stack creation begins and becomes CREATE_COMPLETE after about 7 minutes. Then open SageMakerStudioUrl from the Outputs tab.
The SageMaker Studio AI screen opens, so press Skip.
Open Code Editor and press Open. If it was stopped, start it first and then press Open.
Use CDK for Updates and Parameter Changes
The SageMaker Code Editor screen is now open. Press New Terminal.
Execute the following command in the terminal to clone the latest GenU source code. By cloning the latest source code, GenU version upgrades are possible. For second and subsequent version upgrades where git clone has already been done, please refer to these steps.
git clone https://github.com/aws-samples/generative-ai-use-cases.git
Here's an image showing where to execute commands. As explained in the following image, the Terminal is displayed at the bottom of the screen, so execute commands here.
Open the cloned folder.
Press Yes, I trust the authors.
Open packages/cdk/parameter.ts.
Again, press New Terminal to open the Terminal.
Move to the directory where GenU was cloned.
cd /home/sagemaker-user/generative-ai-use-cases/
Edit the parameter.ts file using the Parameter Store values confirmed in the previous step.
For environments using the default dev, automatic configuration using commands is possible.
Tip
If using environments other than dev, manually copy the Parameter Store values to directly edit the parameter.ts file.
PARAMS=$(aws ssm get-parameter --name "/genu/dev.json" --with-decryption --query "Parameter.Value" --output text)
Next, execute the following command to edit the parameter.ts file.
node -e "
const fs = require('fs');
const params = $PARAMS;
let content = fs.readFileSync('packages/cdk/parameter.ts', 'utf8');
const devStart = content.indexOf('dev: {');
const devEnd = content.indexOf('},', devStart) + 2;
const jsonString = JSON.stringify(params, null, 2);
const indentedJson = jsonString.split('\n').map((line, index) => {
if (index === 0) return line;
return ' ' + line;
}).join('\n');
const newDevSection = 'dev: ' + indentedJson + ',';
content = content.substring(0, devStart) + newDevSection + content.substring(devEnd);
fs.writeFileSync('packages/cdk/parameter.ts', content);
"
The dev file content should be edited as follows:
dev: {
"modelRegion": "us-east-1",
"modelIds": [
"us.anthropic.claude-sonnet-4-20250514-v1:0",
"us.anthropic.claude-opus-4-20250514-v1:0",
"us.deepseek.r1-v1:0",
"us.anthropic.claude-3-7-sonnet-20250219-v1:0",
"us.anthropic.claude-3-5-haiku-20241022-v1:0",
"us.amazon.nova-premier-v1:0",
"us.amazon.nova-pro-v1:0",
"us.amazon.nova-lite-v1:0",
"us.amazon.nova-micro-v1:0"
],
"ragKnowledgeBaseEnabled": true,
"selfSignUpEnabled": false,
"allowedSignUpEmailDomains": [
"gmail.com"
],
"allowedIpV4AddressRanges": null,
"allowedIpV6AddressRanges": null
},
staging: {
// Parameters for staging environment
},
prod: {
// Parameters for production environment
},
This is an example of the screen after editing.
Let's change the parameters. In this procedure, we will change the models to use.
Resolve dependencies.
npm ci
Execution example
sagemaker-user@default:~/generative-ai-use-cases$ npm ci
npm warn deprecated sourcemap-codec@1.4.8: Please use @jridgewell/sourcemap-codec instead
npm warn deprecated rimraf@3.0.2: Rimraf versions prior to v4 are no longer supported
npm warn deprecated inflight@1.0.6: This module is not supported, and leaks memory. Do not use it. Check out lru-cache if you want a good and tested way to coalesce async requests by a key value, which is much more comprehensive and powerful.
npm warn deprecated glob@7.2.3: Glob versions prior to v9 are no longer supported
npm warn deprecated @humanwhocodes/object-schema@2.0.3: Use @eslint/object-schema instead
npm warn deprecated @humanwhocodes/config-array@0.13.0: Use @eslint/config-array instead
npm warn deprecated eslint@8.57.1: This version is no longer supported. Please see https://eslint.org/version-support for other options.
> generative-ai-use-cases@4.2.7 prepare
> husky
added 2416 packages, and audited 2461 packages in 2m
477 packages are looking for funding
run `npm fund` for details
3 moderate severity vulnerabilities
To address all issues (including breaking changes), run:
npm audit fix --force
Run `npm audit` for details.
npm notice
npm notice New minor version of npm available! 11.3.0 -> 11.4.1
npm notice Changelog: https://github.com/npm/cli/releases/tag/v11.4.1
npm notice To update run: npm install -g npm@11.4.1
npm notice
sagemaker-user@default:~/generative-ai-use-cases$
Update GenU. The env=dev parameter specifies the Environment name you want to deploy. The default is dev.
npm run cdk:deploy:quick -- -c env=dev
Execution example
sagemaker-user@default:~/generative-ai-use-cases$ npm run cdk:deploy:quick -- -c env=dev
> generative-ai-use-cases@4.2.7 cdk:deploy:quick
> npm -w packages/cdk run cdk deploy -- --all --asset-parallelism --asset-prebuild=false --concurrency 3 --method=direct --require-approval never --force -c env=dev
> cdk
> cdk deploy --all --asset-parallelism --asset-prebuild=false --concurrency 3 --method=direct --require-approval never --force -c env=dev
`cdk synth` may hang in Docker on Linux 5.6-5.10. See https://github.com/aws/aws-cdk/issues/21379 for workarounds.
Bundling asset GenerativeAiUseCasesStackdev/API/Predict/Code/Stage...
packages/cdk/cdk.out/bundling-temp-d7718a9e133c5da897e2705c0897279a63ba1441a740635cdf58a04513f40f3e-building/index.js 3.3mb ⚠️
⚡ Done in 438ms
added 114 packages, and audited 120 packages in 5s
3 packages are looking for funding
run `npm fund` for details
found 0 vulnerabilities
After some time, deployment is complete. The original GenU has been updated.
GenerativeAiUseCasesStackdev.WebUrl = https://xxxxxxxxx.cloudfront.net
Stack ARN:
arn:aws:cloudformation:us-east-1:xxxxxxxxxx:stack/GenerativeAiUseCasesStackdev/83dbd1f0-3ecb-11f0-a659-0e0290ae6345
✨ Total time: 378.4s
sagemaker-user@default:~/generative-ai-use-cases$
Steps for Updating GenU from the Second Time Onwards
This section introduces the steps for updating GenU from the second time onwards using SageMaker Code Editor. While the first time used git clone, the second time onwards has different steps since it's already cloned.
After opening SageMaker Code Editor, open the GenU directory with Open Folder.
Warning
Open packages/cdk/parameter.ts and note down the values of dev, staging, and prod. Be careful not to forget to take notes, as recovery is difficult if lost!
Move to the git cloned directory.
cd /home/sagemaker-user/generative-ai-use-cases/
Reset all file contents edited in Code Editor to their original state. Please note that if you have edited files other than parameter.ts, they will all be lost.
git reset --hard
Get the latest version source code with the following command:
git pull
Resolve dependencies with npm ci:
npm ci
Edit the parameter.ts file. Restore the previously saved values of dev, staging, and prod.
Update GenU. The env=dev parameter specifies the Environment name you want to deploy. The default is dev.
npm run cdk:deploy:quick -- -c env=dev
The update is now complete!