nantu-io/ntcore-monitoring

NTCore Monitoring: SaaS for ML/AI Model Monitoring

Deployment SOP

1. Create a dedicated cloudwatch log group for customer.
2. Create a dedicated kinesis stream for customer.
3. Update kube/ntcore/monitoring.yml with the cloudwatch log group and kinesis stream name, for example,

   provider:
       type: cloudwatch
       group: /ntcore/monitoring/dev
       ...
   streamProvider:
       type: kinesis
       name: log-event-stream-dev
       ...
   

4. Update kube/ntcore/log-events-processor.yml with the cloudwatch log group and kinesis stream name, for example,

   kclpy.properties: |
       streamName = {kinesis_stream_for_customer}
       applicationName = {application_name_for_customer}
       ...
   cloudwatch.properties: |
       [CloudWatch]
       logGroup = {log_group_for_customer}
       ...
   

5. Update cloud/aws/eksctl.yml with the EKS cluster info, for example,

   metadata:
       name: ntcore
       region: cn-northwest-1
   
   nodeGroups:
       - name: beta
         instanceType: t3.medium
         desiredCapacity: 1
   

6. Apply the kubernetes configs.

Documentation

NTCore documentation: https://nantu-io.github.io/ntcore-doc.

• Quickstart
• Production
• Examples

Community

For Getting Started guides, tutorials, and API reference check out our docs.

To report a bug, file a documentation issue, or submit a feature request, please open a GitHub issue.

License

NTCore is licensed under Apache 2.0.