Skip to content
This repository was archived by the owner on Jul 11, 2025. It is now read-only.

Commit 507f1f9

Browse files
committed
Claude updates
1 parent 00f8670 commit 507f1f9

File tree

3 files changed

+162
-25
lines changed

3 files changed

+162
-25
lines changed

deployment/aws.md renamed to deployment/aws-ec2.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -56,4 +56,4 @@ Coming soon - detailed content!
5656

5757
## Troubleshooting
5858

59-
Coming soon - detailed content!
59+
Coming soon - detailed content!

deployment/aws-rke.md

Lines changed: 117 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,117 @@
1+
---
2+
title: Amazon Web Services (RKE)
3+
layout: default
4+
nav_order: 4
5+
parent: Deployment
6+
grand_parent: TrustGraph Documentation
7+
---
8+
9+
# Amazon Web Services (RKE) Deployment
10+
11+
Deploy TrustGraph on Amazon Web Services using RKE2 Kubernetes and AWS Bedrock integration.
12+
13+
## Overview
14+
15+
TrustGraph provides a complete AWS deployment solution using **Pulumi** (Infrastructure as Code) that automatically provisions and configures an RKE2 Kubernetes cluster with AWS Bedrock LLM integration for a production-ready TrustGraph deployment.
16+
17+
## What You Get
18+
19+
The AWS deployment includes:
20+
21+
- **RKE2 Kubernetes cluster** with 1 server and multiple agent nodes
22+
- **IAM roles and policies** with Bedrock access permissions
23+
- **EBS CSI driver** for persistent storage provisioning
24+
- **Complete TrustGraph stack** deployed and configured
25+
- **AWS Bedrock integration** with Claude 3.5 Haiku (configurable)
26+
- **Mistral Nemo Instruct** endpoint configured
27+
- **Secrets management** for secure credential handling
28+
- **Monitoring and observability** with Grafana
29+
- **Web workbench** for document processing and Graph RAG
30+
31+
## Deployment Method
32+
33+
The deployment uses **Pulumi**, an Infrastructure as Code tool that:
34+
35+
- Has an open-source license
36+
- Uses general-purpose programming languages (TypeScript/JavaScript)
37+
- Provides testable infrastructure code
38+
- Offers retryable deployments
39+
- Supports local or S3 state management
40+
41+
## Architecture
42+
43+
**Kubernetes Platform**: RKE2 (Rancher Kubernetes Engine 2)
44+
**Node Configuration**: t3a.xlarge instances (configurable)
45+
**Default Setup**: 3 nodes, Amazon Linux 2023
46+
**Network**: Custom VPC with configurable CIDR blocks
47+
**Storage**: EBS volumes with automatic provisioning
48+
**LLM Service**: AWS Bedrock with multiple model options
49+
50+
## Quick Process Overview
51+
52+
1. **Install Pulumi** and dependencies
53+
2. **Configure AWS credentials** via `.aws/credentials`
54+
3. **Set up Bedrock model access** in AWS console
55+
4. **Customize configuration** in `Pulumi.STACKNAME.yaml`
56+
5. **Deploy** with `pulumi up`
57+
6. **Wait for initialization** (~30 seconds after pods are running)
58+
7. **Access services** via port-forwarding
59+
60+
## Configuration Options
61+
62+
Customizable settings include:
63+
64+
- **Environment**: dev, staging, production
65+
- **Region**: us-west-2 (default, configurable)
66+
- **VPC CIDR**: 172.38.0.0/16
67+
- **Node type**: t3a.xlarge (configurable)
68+
- **Node count**: 3 (configurable)
69+
- **AMI**: Amazon Linux 2023
70+
71+
## Access Points
72+
73+
Once deployed, you'll have access to:
74+
75+
- **TrustGraph API**: Port 8088
76+
- **Web Workbench**: Port 8888 (document processing, Graph RAG)
77+
- **Grafana Monitoring**: Port 3000
78+
79+
## Bedrock Integration
80+
81+
The deployment includes AWS Bedrock integration with:
82+
83+
- **Default Model**: Claude 3.5 Haiku
84+
- **Alternative Models**: Mistral Nemo Instruct and others
85+
- **Model Access**: Requires enabling specific models in Bedrock console
86+
- **Secure Access**: IAM roles and policies automatically configured
87+
88+
## Complete Documentation
89+
90+
For detailed step-by-step instructions, configuration options, and troubleshooting, visit:
91+
92+
**[TrustGraph AWS RKE Deployment Guide](https://github.com/trustgraph-ai/pulumi-trustgraph-aws-rke)**
93+
94+
The repository contains:
95+
- Complete Pulumi deployment code
96+
- RKE2 cluster configuration
97+
- AWS Bedrock integration setup
98+
- Detailed setup instructions
99+
- Troubleshooting guides
100+
- Customization options
101+
102+
## Important Notes
103+
104+
**Storage Cleanup**: The EBS CSI driver doesn't automatically clean up volumes on teardown. Manual cleanup may be required in the EC2 console.
105+
106+
**Initialization Time**: Allow ~30 seconds after all pods are running for application initialization to complete.
107+
108+
**Bedrock Access**: Ensure you have enabled the required models in your AWS Bedrock console before deployment.
109+
110+
## Next Steps
111+
112+
After deployment, you can:
113+
- Load documents through the web workbench
114+
- Test Graph RAG queries with Bedrock models
115+
- Monitor processing through Grafana
116+
- Scale the cluster by adjusting node count
117+
- Integrate with other AWS services

deployment/gcp.md

Lines changed: 44 additions & 24 deletions
Original file line numberDiff line numberDiff line change
@@ -12,44 +12,64 @@ Deploy TrustGraph on Google Cloud Platform using Google Kubernetes Engine (GKE)
1212

1313
## Overview
1414

15-
GCP provides a robust, scalable platform for deploying TrustGraph in production environments with enterprise-grade features.
15+
TrustGraph provides a complete GCP deployment solution using **Pulumi** (Infrastructure as Code) that automatically provisions and configures all necessary GCP resources for a production-ready TrustGraph deployment.
1616

17-
## Prerequisites
17+
## What You Get
1818

19-
Coming soon - detailed content!
19+
The GCP deployment includes:
2020

21-
## GKE Cluster Setup
21+
- **GKE Kubernetes cluster** with 2-node pool
22+
- **Service accounts** with VertexAI access
23+
- **Complete TrustGraph stack** deployed and configured
24+
- **VertexAI Gemini Flash 1.5** LLM integration
25+
- **Secrets management** for secure configuration
26+
- **Monitoring and observability** with Grafana
27+
- **Web workbench** for document processing and Graph RAG
2228

23-
Coming soon - detailed content!
29+
## Deployment Method
2430

25-
## Container Registry
31+
The deployment uses **Pulumi**, an Infrastructure as Code tool that:
2632

27-
Coming soon - detailed content!
33+
- Has an open-source license
34+
- Uses general-purpose programming languages (TypeScript/JavaScript)
35+
- Provides testable infrastructure code
36+
- Offers retryable deployments
37+
- Supports local or cloud state management
2838

29-
## Deployment Configuration
39+
## Quick Process Overview
3040

31-
Coming soon - detailed content!
41+
1. **Install Pulumi** and dependencies
42+
2. **Configure GCP credentials** using `gcloud auth login`
43+
3. **Customize configuration** in `Pulumi.STACKNAME.yaml`
44+
4. **Deploy** with `pulumi up`
45+
5. **Access services** via port-forwarding or load balancers
3246

33-
## Persistent Storage
47+
## Access Points
3448

35-
Coming soon - detailed content!
49+
Once deployed, you'll have access to:
3650

37-
## Networking & Security
51+
- **TrustGraph API**: Port 8088
52+
- **Web Workbench**: Port 8888 (document processing, Graph RAG)
53+
- **Grafana Monitoring**: Port 3000
3854

39-
Coming soon - detailed content!
55+
## Complete Documentation
4056

41-
## Monitoring & Logging
57+
For detailed step-by-step instructions, configuration options, and troubleshooting, visit:
4258

43-
Coming soon - detailed content!
59+
**[TrustGraph GCP Deployment Guide](https://github.com/trustgraph-ai/pulumi-trustgraph-gke)**
4460

45-
## Scaling
61+
The repository contains:
62+
- Complete Pulumi deployment code
63+
- Configuration templates
64+
- Detailed setup instructions
65+
- Troubleshooting guides
66+
- Customization options
4667

47-
Coming soon - detailed content!
68+
## Next Steps
4869

49-
## Cost Optimization
50-
51-
Coming soon - detailed content!
52-
53-
## Troubleshooting
54-
55-
Coming soon - detailed content!
70+
After deployment, you can:
71+
- Load documents through the web workbench
72+
- Test Graph RAG queries
73+
- Monitor processing through Grafana
74+
- Scale the cluster as needed
75+
- Integrate with existing GCP services

0 commit comments

Comments
 (0)