Wed, Oct 08, 25, DETAILED RESTORATION LOG - Auto-imported from uconGPT project

Auto-imported from: D:/repos/aiegoo/uconGPT/eng2Fix/kor2fix/DETAILED_RESTORATION_LOG.md
Original filename: DETAILED_RESTORATION_LOG.md
Import date: Wed, Oct 08, 25

Docker Desktop and WSL Restoration - Detailed Step-by-Step Record

Date: October 6, 2025
Duration: ~3 hours
Operator: aiegoo + GitHub Copilot
Target: Restore Korean language processing platform from August 25, 2025 Linked Issues: #31, #29, #26

📋 Initial Situation Assessment

Environment Analysis

  • OS: Windows 11
  • Current Directory: d:\repos\aiegoo\uconGPT\eng2Fix\kor2fix
  • Branch: environment-setup-v1
  • Issue: Python path error: bash: C:\Users\hsyyu\anaconda3/bin/python: No such file or directory
  • Goal: Restore Docker Desktop to run on D drive and recover 18 containers from August 25, 2025

Original Container Inventory (from screenshot)

18 containers were running on August 25, 2025:

  • mongodb, kor2unity-offline, korean-frontend-offline, korean-monitoring-offline
  • korean-learning-api-offline, korean-audio-api-offline, korean-llm-api-offline
  • korean-redis-offline, rasa-actions-offline, korean-file-server-offline
  • korean-mongodb-offline, my-korean-platform, korean-mongodb
  • korean-rasa-actions, korean-llm-api, korean-redis
  • korean-file-server-minimal, voice-upload-server-latest

🔧 Step-by-Step Restoration Process

Phase 1: Docker Desktop Installation (10:00 AM - 11:00 AM)

Step 1.1: Verify Docker Desktop Installer

# Command executed:
Test-Path "C:\Users\hsyyu\Downloads\Docker Desktop Installer.exe"
# Result: True - installer confirmed present

Step 1.2: Install Docker Desktop with Admin Privileges

# Command executed:
Start-Process -FilePath "C:\Users\hsyyu\Downloads\Docker Desktop Installer.exe" -Verb RunAs -ArgumentList "install --quiet --accept-license"
# Result: Docker Desktop 4.47.0 installed successfully

Step 1.3: Monitor Installation Progress

# Commands executed:
Get-Process | Where-Object {$_.ProcessName -like "*Docker*"}
Start-Sleep -Seconds 30
Test-Path "C:\Program Files\Docker\Docker\Docker Desktop.exe"
# Result: Installation completed, executable confirmed

Phase 2: D Drive Storage Configuration (11:00 AM - 12:00 PM)

Step 2.1: Create D Drive Directory Structure

# Commands executed:
New-Item -ItemType Directory -Path "D:\Docker" -Force
New-Item -ItemType Directory -Path "D:\Docker\data" -Force
New-Item -ItemType Directory -Path "D:\Docker\data\vms" -Force

Step 2.2: Configure Docker Desktop Settings

File: C:\Users\hsyyu\AppData\Roaming\Docker\settings.json

{
  "settingsVersion": 1,
  "dataFolder": "D:\\Docker\\data",
  "enableAnalytics": false,
  "autoStart": true,
  "displayedTutorial": true,
  "wslEngineEnabled": true,
  "useWindowsContainers": false,
  "hyperVCustomVmPath": "D:\\Docker\\data\\vms"
}

Step 2.3: Configure Docker Daemon for D Drive

File: C:\Users\hsyyu\.docker\daemon.json

{
  "builder": {
    "gc": {
      "defaultKeepStorage": "20GB",
      "enabled": true
    }
  },
  "experimental": false,
  "data-root": "D:\\Docker\\data\\docker"
}

Phase 3: WSL Distribution Migration (12:00 PM - 1:00 PM)

Step 3.1: Check WSL Status

# Command executed:
wsl --list --verbose
# Result: Ubuntu-22.04 Running, docker-desktop detected

Step 3.2: Export Docker Desktop WSL Distribution

# Command executed:
wsl --export docker-desktop "D:\Docker\data\docker-desktop-backup.tar"
# Result: Export completed successfully (48 MB)

Step 3.3: Unregister and Re-import to D Drive

# Commands executed:
wsl --unregister docker-desktop
New-Item -ItemType Directory -Path "D:\Docker\data\wsl\docker-desktop" -Force
wsl --import docker-desktop "D:\Docker\data\wsl\docker-desktop" "D:\Docker\data\docker-desktop-backup.tar"
# Result: WSL distribution successfully moved to D drive

Phase 4: JSON/UTF-8 BOM Issues Resolution (1:00 PM - 1:30 PM)

Step 4.1: Fix JSON Parsing Error

Error encountered: invalid character 'ï' looking for beginning of value Root cause: UTF-8 BOM in JSON files

Step 4.2: Recreate daemon.json with Proper Encoding

# Commands executed:
Remove-Item "C:\Users\hsyyu\.docker\daemon.json" -Force
$daemonConfig = @'
{
  "builder": {
    "gc": {
      "defaultKeepStorage": "20GB",
      "enabled": true
    }
  },
  "experimental": false
}
'@
[System.IO.File]::WriteAllText("C:\Users\hsyyu\.docker\daemon.json", $daemonConfig, [System.Text.UTF8Encoding]::new($false))

Step 4.3: Fix Docker Config Credential Issues

# Commands executed:
Remove-Item "$env:USERPROFILE\.docker\config.json" -Force
$config = '{"auths": {}, "credsStore": ""}'
[System.IO.File]::WriteAllText("$env:USERPROFILE\.docker\config.json", $config, [System.Text.UTF8Encoding]::new($false))

Phase 5: Container Images Recovery (1:30 PM - 2:30 PM)

Step 5.1: Identify Docker Hub Repositories

Research: Found aiegoo Docker Hub account with matching repositories:

  • aiegoo/voice-upload-server (786MB)
  • aiegoo/llm-api (7.41GB)
  • aiegoo/audio-api (9.82GB)
  • aiegoo/korean-api (7.33GB)
  • aiegoo/rasa-actions (248MB)
  • aiegoo/monitoring

Step 5.2: Pull Container Images from Docker Hub

# Commands executed:
docker pull aiegoo/voice-upload-server
docker pull aiegoo/llm-api
docker pull aiegoo/audio-api
docker pull aiegoo/korean-api
docker pull aiegoo/rasa-actions
docker pull aiegoo/monitoring
docker pull mongo:5.0
docker pull mongo:7.0
docker pull redis:7-alpine

Results:

REPOSITORY                   TAG       IMAGE ID       CREATED       SIZE
aiegoo/voice-upload-server   latest    ea8d4001b3a7   6 weeks ago   786MB
aiegoo/audio-api             latest    80bcc84aa1f0   7 weeks ago   9.82GB
aiegoo/korean-api            latest    3f3e8318b06c   7 weeks ago   7.33GB
aiegoo/llm-api               latest    e8747a5978e1   7 weeks ago   7.41GB
aiegoo/rasa-actions          latest    4ef9d8b6666e   7 weeks ago   248MB
aiegoo/monitoring            latest    [ID]            7 weeks ago   [SIZE]

Phase 6: Docker Compose Configuration (2:30 PM - 3:00 PM)

Step 6.1: Create Comprehensive Docker Compose File

File: D:\Docker\data\korean-platform-docker-compose.yml

  • Configured 15+ services matching original container names
  • Set up D drive volume mappings
  • Configured service networking and dependencies
  • Port mappings: 8200-8212 range

Step 6.2: Create Volume Directory Structure

# Commands executed:
$volumeDirs = @(
    "D:\Docker\data\volumes\mongodb_data",
    "D:\Docker\data\volumes\korean_mongodb_data", 
    "D:\Docker\data\volumes\redis_data",
    "D:\Docker\data\volumes\voice_uploads",
    "D:\Docker\data\volumes\frontend_static",
    "D:\Docker\data\volumes\file_server_data"
)
foreach ($dir in $volumeDirs) { New-Item -ItemType Directory -Path $dir -Force }

Phase 7: Service Restoration Testing (3:00 PM - 3:30 PM)

Step 7.1: Start Core Services

# Commands executed:
cd D:\Docker\data
docker compose -f korean-platform-docker-compose.yml up -d mongodb korean-redis

Step 7.2: Verify Service Status

# Command executed:
docker ps
# Results:
CONTAINER ID   IMAGE            PORTS                     NAMES
4243d3d71fd2   redis:7-alpine   0.0.0.0:8210->6379/tcp    korean-redis
108e413f1c39   mongo:5.0        0.0.0.0:8202->27017/tcp   mongodb

Phase 8: Git Repository Documentation (3:30 PM - 4:00 PM)

Step 8.1: Create Detailed Documentation

File: DOCKER_RESTORATION_PROGRESS.md

  • Complete restoration timeline
  • Technical configuration details
  • Container inventory mapping
  • Issue resolution log

Step 8.2: Git Operations and Milestone Tagging

# Authentication setup (as aiegoo user, not root):
git config --global user.name "aiegoo"
git config --global user.email "[email protected]"
git config --global core.editor vim

# Add documentation:
git add DOCKER_RESTORATION_PROGRESS.md

# Commit with detailed message:
git commit -m "feat: Docker Desktop and WSL restoration milestone

- Successfully installed Docker Desktop 4.47.0 on D drive
- Migrated docker-desktop WSL distribution to D:/Docker/data 
- Recovered container images from Docker Hub (aiegoo/* repositories)
- Restored MongoDB and Redis services
- Created comprehensive Docker Compose configuration
- Fixed UTF-8 BOM and credential issues
- Documented complete restoration progress

Milestone: Docker-WSL-Restoration-v1.0
Restores: Korean language processing platform from Aug 25, 2025
Next: API services restoration and frontend rebuilding"

# Create milestone tag:
git tag -a 'docker-wsl-restoration-v1.0' -m 'Docker Desktop and WSL Restoration Milestone v1.0'

# Handle divergent branches (remote had updates):
git pull --no-rebase origin environment-setup-v1
# Merged 74 files, 7265 insertions from remote

# Push all changes:
git push origin environment-setup-v1
git push origin docker-wsl-restoration-v1.0

Phase 9: Terminal & Editor Experience Restoration (4:00 PM - 6:00 PM)

Step 9.1: Replace Broken Oh My Posh Prompt

  • Observed garbled glyphs/icons in existing Oh My Posh prompt on Windows Terminal
  • Requirement pivot from visuals to readability: GitKraken-style abbreviations, collapsible status line, and zero font dependencies
  • Decision: implement custom Bash prompt while keeping optional powerline support

Step 9.2: Deploy Compact GitKraken-Style Prompt

  • Source file: D:\Docker\data\terminal_configs\compact_bashrc.sh
  • Highlights:
    • build_prompt() renders a single-line status with smart path truncation and git glyphs (● ± … ↑ ↓)
    • Shortcut banner on shell start: ⚡ Korean Dev Ready | Shortcuts: ks kd kt ka | Git: g gst ga gc gp
    • Git shortcuts (g, gst, ga, gc, gp, gpl, etc.) and Korean dev helpers (ks, kd, kt, ka)
    • Fallback prompt when git repository metadata is unavailable

  • Synced into WSL user profile:

    cp /mnt/d/Docker/data/terminal_configs/compact_bashrc.sh ~/.bashrc
sed -i 's/\r$//' ~/.bashrc

Step 9.3: Add Powerline with Safe Fallback

  • Installed bindings: pip install --user powerline-status

  • Added conditional sourcing inside .bashrc:

    POWERLINE_BINDING="$HOME/.local/lib/python3.13/site-packages/powerline/bindings/bash/powerline.sh"
command -v powerline-daemon >/dev/null 2>&1 && powerline-daemon -q
  • Custom shell theme placed at ~/.config/powerline/themes/shell/default.json (env, cwd, git status segments)
  • Behavior: powerline segments appear when patched Nerd Font is active; otherwise prompt falls back to ASCII-only separators

Step 9.4: Modernize Vim Workflow

  • Source file: D:\Docker\data\terminal_configs\compact_vimrc.vim → synced to ~/.vimrc
  • Migrated from Pathogen/Vundle to vim-plug
  • Plugins installed via vim +PlugInstall +qall:
    • vim-airline/vim-airline + vim-airline-themes
    • preservim/nerdtree
    • tpope/vim-fugitive
    • mhinz/vim-startify
    • airblade/vim-gitgutter
    • christoomey/vim-tmux-navigator
    • morhetz/gruvbox
  • Startify configured with sessions, recent kor2fix files, and project bookmarks
  • ASCII-friendly airline symbols provided when AIRLINE_POWERLINE_FONTS is unset (per vi.stackexchange.com/a/3361)
  • Auto-creates ~/.vim/undo and ~/.vim/session directories for persistence

Step 9.5: Validation

  • Powerline prompt renders as aiegoo  /  mnt  d  when fonts available; otherwise /mnt/d ›
  • gst alias shows compact git status; Korean helper commands reachable via ks, kd, kt, ka
  • Launching vim now opens Startify dashboard with tmux navigation shortcuts enabled

Phase 10: Core Service Orchestration (6:00 PM - 8:00 PM)

Step 10.1: Rebuild Docker Compose Topology

  • Replaced legacy compose stack with branch-aligned services (backend, frontend, rasa-core, rasa-actions, mongodb, redis, health-monitor)
  • Added optional profiles for ollama (LLM) and qdrant (vector DB) to defer heavy downloads (docker compose --profile llm up -d)
  • Created .env from template to centralize dedicated port mappings and backend URLs

Step 10.2: Container Image Refresh

cd D:\repos\aiegoo\uconGPT\eng2Fix\kor2fix
& "C:\Program Files\Docker\Docker\resources\bin\docker.exe" compose build backend frontend rasa-actions rasa-core
  • Added Python package initializers (backend/__init__.py, backend/api/__init__.py) so Uvicorn can import backend.api.main
  • Simplified frontend delivery via nginx:1.27-alpine serving existing HTML templates
  • Hardened Rasa build pipeline with a minimal Korean NLU dataset and automated training during image build

Step 10.3: Resolve Build & Runtime Issues

  • Installed action server dependencies as root inside the image to avoid pip permission errors, then dropped back to the default user
  • Declared action_ask_llm in domain.yml and added fallback rules to unblock rasa train
  • Removed the obsolete Compose version key to silence v2 CLI warnings

Step 10.4: Launch Core Services on Dedicated Ports

# Free conflicting legacy Mongo container
& "C:\Program Files\Docker\Docker\resources\bin\docker.exe" stop mongodb
& "C:\Program Files\Docker\Docker\resources\bin\docker.exe" rm mongodb

# Start branch services
& "C:\Program Files\Docker\Docker\resources\bin\docker.exe" compose up -d backend frontend rasa-actions rasa-core mongodb redis health-monitor

# Verify health endpoints
Invoke-WebRequest http://localhost:8201/health | Select Content
Invoke-WebRequest http://localhost:8204/status | Select StatusCode
  • docker compose ps shows all core containers healthy on ports 8201-8209 and 8212
  • Confirmed FastAPI /api/llm/chat placeholder echo and Rasa REST channel are reachable for integration tests

🎯 Current Status Summary

✅ Successfully Completed

  1. Docker Desktop: Installed and configured on D drive
  2. WSL Migration: docker-desktop distribution moved to D:\Docker\data\wsl
  3. Container Images: 7 key images recovered from Docker Hub
  4. Core Services: MongoDB (port 8202) and Redis (port 8210) operational
  5. Storage Configuration: All Docker data routed to D drive
  6. Documentation: Complete step-by-step record created
  7. Version Control: Changes committed and tagged as milestone
  8. Terminal & Editor: GitKraken-style prompt, powerline-ready pipeline, and vim workflow restored

📋 Image Recovery Status

  • ✅ aiegoo/voice-upload-server:latest (786MB)
  • ✅ aiegoo/llm-api:latest (7.41GB)
  • ✅ aiegoo/audio-api:latest (9.82GB)
  • ✅ aiegoo/korean-api:latest (7.33GB)
  • ✅ aiegoo/rasa-actions:latest (248MB)
  • ✅ aiegoo/monitoring:latest
  • ✅ mongo:5.0, mongo:7.0, redis:7-alpine

🚧 Pending Restoration (13 containers remaining)

  • korean-frontend-offline (needs rebuild)
  • kor2unity-offline (main service)
  • korean-learning-api-offline (image available)
  • korean-audio-api-offline (image available)
  • korean-llm-api-offline (image available)
  • korean-file-server-offline, korean-file-server-minimal
  • korean-monitoring-offline (image available)
  • rasa-actions-offline, korean-rasa-actions
  • korean-mongodb-offline (DB setup needed)
  • my-korean-platform (integration service)

📁 Final Directory Structure

D:\Docker\
├── data\
│   ├── docker\                    # Docker daemon data
│   ├── vms\                      # Hyper-V VMs  
│   ├── wsl\                      # WSL distributions
│   │   └── docker-desktop\       # Migrated WSL distro
│   ├── volumes\                  # Named volumes
│   │   ├── mongodb_data\
│   │   ├── korean_mongodb_data\
│   │   ├── redis_data\
│   │   ├── voice_uploads\
│   │   ├── frontend_static\
│   │   └── file_server_data\
│   ├── korean-platform-docker-compose.yml
│   └── docker-desktop-backup.tar
└── user-config\                  # Configuration backups
    └── daemon.json

🔧 Technical Configuration Details

Docker Desktop Settings

  • Data Folder: D:\Docker\data
  • WSL Engine: Enabled
  • Hyper-V VMs: D:\Docker\data\vms
  • Settings Version: 1

WSL Configuration

  • Distribution: docker-desktop (WSL2)
  • Location: D:\Docker\data\wsl\docker-desktop
  • Size: 48MB (exported)
  • User Setup: aiegoo user with SSH keys

Network Configuration

  • Bridge Network: korean-network (172.20.0.0/16)
  • Port Range: 8200-8212 (dedicated to Korean platform)
  • Service Discovery: Docker internal DNS

🚀 Next Phase Requirements

Immediate Actions Needed

  1. Start remaining API services using existing images
  2. Investigate frontend rebuild requirements
  3. Configure Rasa integration with proper models
  4. Set up file server with persistent storage
  5. Implement health monitoring across all services

Success Metrics

  • All 18 original containers running
  • Services accessible on designated ports
  • Data persistence on D drive confirmed
  • Inter-service communication functional
  • Platform ready for Korean language processing

Restoration Milestone: docker-wsl-restoration-v1.0
Git Tag: docker-wsl-restoration-v1.0
Commit: df595e5 on environment-setup-v1 branch
Next Phase: API Services Restoration (Issue #TBD)