Provide three tiers of the install script for both 24.04 and 26.04:
- Originals (ubuntu-post-install-24.04.sh / -26.04.sh): restored to their
true original state, with Keycloak intact, as a fallback baseline. (This
reverts the in-place Keycloak removal from the previous commit; the cleanup
now lives in the -no-keycloak variants instead.)
- -no-keycloak.sh: Keycloak fully removed, Authelia as the SSO/2FA option.
- -crowdsec.sh: builds on -no-keycloak and replaces fail2ban entirely with
CrowdSec (SSH via auth.log/sshd collection, Caddy via caddy collection + log
acquisition, firewall bouncer for enforcement, plus geo-blocking and
community IP-reputation blocklists).
Add SCRIPT-VARIANTS.md documenting the three tiers and how the Authelia /
fail2ban / CrowdSec security layers differ.
All variants pass 'bash -n'.
https://claude.ai/code/session_017eA2qqq9jfF2tNtpUYL8vK
Keycloak never reliably ran (fiddly reverse-proxy/hostname config) and the
repo has standardized on Authelia for SSO + 2FA. This rips Keycloak out
entirely:
- Delete the install block, whiptail menu entry, uninstall plumbing, and
EXISTING_SERVICES detection from both 24.04 and 26.04 scripts
- Delete docker-compose-keycloak.yml, fix-keycloak-proxy.sh, and
KEYCLOAK-SETUP-GUIDE.md
- Remove the Keycloak block from caddy-setup-helper.sh
- Update docs (CADDY-FAIL2BAN-SETUP.md, SECURITY-IMPROVEMENTS.md,
NEW-SCRIPT-STRUCTURE.md, SCRIPT-FLOW-INTRO.txt) to reference Authelia
Also documents the fail2ban/Authelia overlap: Authelia handles failed-login
regulation (per-account lockout); the Caddy fail2ban jail is complementary
firewall-level IP banning. Neither does geo-blocking — noted CrowdSec / Caddy
GeoIP as the path for that.
https://claude.ai/code/session_017eA2qqq9jfF2tNtpUYL8vK
Authelia's installer auto-injects the (authelia) snippet and auth portal
block into the Caddyfile. Moving Caddy first means that injection works
in a single fresh install run without manual follow-up.
https://claude.ai/code/session_01FvqXSZyk3g7rUombwLcprZ
The working authelia-setup repo uses the Authelia 4.38+ password
hashing config format with a nested argon2 block and variant key.
The previous version used the old flat format which is rejected by
Authelia 4.39.20 validation. Fix both 24.04 and 26.04 scripts.
Before (broken):
password:
algorithm: argon2id
iterations: 3
...
After (correct for 4.38+):
password:
algorithm: argon2
argon2:
variant: argon2id
iterations: 3
...
https://claude.ai/code/session_01FvqXSZyk3g7rUombwLcprZ
- Rename ubuntu-post-install.sh → ubuntu-post-install-24.04.sh to match 26.04 naming convention
- Add --allow-server-ssh systemd override to 24.04 NetBird install (already in 26.04)
- Add full Authelia install block to 26.04 (matching what was added to 24.04)
- Whiptail menu, uninstall, detection, default vars, parse flags, install block
- Both scripts now identical in Authelia and NetBird SSH behavior
https://claude.ai/code/session_01FvqXSZyk3g7rUombwLcprZ
Ports the full working configuration from outis1one/authelia-setup:
- Authelia 4.39.20 on caddy_net with secrets via env var files
- Generates jwt/session/storage secrets with openssl at install time
- Prompts for domain, admin user, SMTP settings, timezone
- Generates argon2id password hash via Docker during install
- Writes configuration.yml, users.yml, docker-compose.yml, .env
- Auto-injects (authelia) snippet + auth portal block into Caddyfile
- Creates caddy_net Docker network if missing
- Adds Authelia to whiptail service menu, uninstall list, and detection
- Adds Authelia snippet as commented example in new Caddyfile template
https://claude.ai/code/session_01FvqXSZyk3g7rUombwLcprZ
Without --allow-server-ssh, NetBird prompts for re-authentication on
every SSH connection. This adds a systemd drop-in override at
/etc/systemd/system/netbird.service.d/ssh-server.conf so the flag
is set automatically on every boot without manual intervention.
https://claude.ai/code/session_017jFG5YuHf2CCGS5HiheoeM
NetBird v0.60.0 removed the built-in SSH server ('netbird ssh <peer-name>').
SSH now routes through standard openssh-server via a drop-in config at
/etc/ssh/sshd_config.d/99-netbird.conf on port 22022.
- Ensure openssh-server is installed and enabled when NetBird is selected
- Remove all 'netbird ssh <peer-name>' references
- Update instructions: connect via 'ssh user@<netbird-ip>' using netbird status
- Note the dashboard step: Peers > [peer] > SSH to enable per-peer SSH access
- Applied across install section, SSH summary, and next-steps section
https://claude.ai/code/session_017jFG5YuHf2CCGS5HiheoeM
Based on the 24.04 script with the following updates:
- Updated all version references from 24.04 to 26.04
- Fixed Tailscale APT repo from hardcoded 'jammy' to 'resolute' codename
(was a latent bug even on 24.04; jammy is 22.04)
- All other repos (Docker, Kopia, NodeSource) already use dynamic
VERSION_CODENAME detection and work as-is on 26.04
https://claude.ai/code/session_017jFG5YuHf2CCGS5HiheoeM
The Immich CLI uses the File global class which requires Node.js v20+.
The script previously fell back to apt install nodejs which gives v18 on
Ubuntu and fails with "ReferenceError: File is not defined".
Now checks the Node.js major version first. If < 20, offers to install
Node.js 22 LTS from NodeSource before proceeding.
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
- Update database image from deprecated tensorchord/pgvecto-rs:pg14-v0.2.0
to ghcr.io/immich-app/postgres:14-vectorchord0.4.3-pgvectors0.2.0
- Remove database command block (vectors.so) that blocks VectorChord from
loading — the new image handles shared_preload_libraries internally
- Remove bogus healthcheck using non-existent googlechecksum function
- Create required subdirectories (thumbs, upload, backups, library, profile,
encoded-video) with .immich marker files before first start — fixes
ENOENT crash on encoded-video/.immich
- Update Valkey from 8-bookworm to 9-bookworm
- Fix status output: show UPLOAD_LOCATION instead of unset PHOTOS_DIR
for external library strategy
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
The healthcheck test value contained "googlechecksum: $$Chksum"
which YAML interprets as a mapping separator (colon-space in an
unquoted scalar). This caused "mapping values are not allowed in
this context" on the command line below it.
Fixed by using >- block scalar for the healthcheck test and
multi-line array format for the postgres command, matching the
official Immich docker-compose format.
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
Three issues fixed:
1. SSH/VPN/Remote Desktop rerun detection: When services are already
configured, ask a single "Reconfigure?" question instead of
prompting through every individual service again.
2. Whiptail selection respected: When user selects specific services
in the whiptail menu, skip all unrelated sections (linux-to-sync,
Caddy Legacy, Kopia, local backup, cloud backup, UFW) instead of
prompting for each one.
3. Import-photos.sh reminder: Clarify that import-photos.sh should
be run AFTER the main setup script completes, not during.
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
- apt upgrade at end of script now prompts (default: no) instead of
running unconditionally. This prevents unwanted package upgrades
when running the script just to install a new service.
- Immich setup: ask "existing photos?" first (default: no) so fresh
installs only get one path question. Strategy choice and path prompts
now only appear when relevant to the chosen strategy.
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
The user should decide *what* they want (import vs external library)
before being asked *where* their photos are. The "what" frames the
context for the "where" question.
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
The import-photos.sh script creates the admin account via API, so
"Open browser and create admin account" is no longer needed for the
import path. Post-setup is now just:
Import path (Strategy 1 + existing photos):
"Run ~/docker/immich/import-photos.sh" — one step, no browser
External library path (Strategy 2):
Still requires web UI for library creation (3 steps)
No existing photos:
Still requires web UI for storage template (2 steps)
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
The import-photos.sh helper script now handles the entire setup flow
without requiring any manual web UI interaction:
1. Checks if Immich is initialized via /api/server/config
2. If uninitialized: creates admin account via /api/auth/admin-sign-up
(prompts for email, password, name)
3. If already initialized: prompts for existing credentials
4. Logs in via /api/auth/login to get bearer token
5. Creates API key via /api/api-keys automatically
6. Configures storage template via /api/system-config (PUT)
7. Installs immich-cli (checks immich → npx → npm → offers apt install)
8. Runs recursive upload with EXIF date preservation
Also accepts an API key as argument to skip account setup for re-runs.
Fixes:
- Removed -f flag from curl calls that suppressed error details
- Fixed $? check after login (now uses if ! command pattern)
- Variable naming collision (INSTALL_NODE → INSTALL_NODE_YN)
- Post-setup instructions now just say "run the import script"
instead of listing manual steps
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
When Strategy 1 (unified library) is chosen with existing photos, the
setup now generates an import-photos.sh script that automates the entire
import process:
- Verifies Immich is running (API health check)
- Prompts for API key with validation
- Configures the storage template via API automatically (uses python3
for JSON manipulation, falls back to manual instructions)
- Installs immich-cli via npx if Node.js available, offers to install
Node.js if not
- Runs the upload with --recursive from the baked-in source path
- Shows photo count and progress
Also:
- Ask where existing photos currently live (separate from library path)
- Use that path for both the external library mount (Strategy 2) and
the import script source (Strategy 1)
- Post-setup instructions reduced to 3 steps: create account, get API
key, run import script
- .env comments reference the import script instead of raw CLI commands
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
Replace the confusing two-prompt (upload/external) configuration with a
cohesive strategy-based flow:
- One question for photo library path instead of two separate prompts
- Ask if user has existing photos, then present two clear strategies:
[1] Import everything into Immich (unified library, recommended)
[2] Keep existing photos in place (external library, read-only)
- Generate strategy-specific docker-compose.yml (no unused external
mount when not needed)
- Strategy-specific .env files with relevant instructions only
- Fix nesting issue: external library uploads now go to sibling dir
instead of subfolder (prevents duplicate scan)
- Unified post-setup instructions that match chosen strategy
https://claude.ai/code/session_01NAtxAkC5t6YVb3gcP1VcX8
- Check for existing .env file instead of extracting from docker-compose
- Replace KC_PROXY with KC_PROXY_HEADERS in .env
- Add KC_PROXY_HEADERS if missing
- Remove KC_PROXY from docker-compose.yml if present
- Show current configuration before restart
- Handles both migration scenarios:
1. Old config with KC_PROXY in .env
2. Existing .env without proxy settings
- Replace deprecated KC_PROXY=edge with KC_PROXY_HEADERS=xforwarded
- Fixes 'Hostname v1 options [proxy] are still in use' warning
- Convert docker-compose-keycloak.yml to use .env file
- Remove hardcoded passwords from docker-compose.yml
- Add comprehensive .env template in comments
- Update deployment instructions and production checklist
- Resolves CORS and secure context warnings
- All credentials now in .env with proper security
This comprehensive update addresses multiple security and usability issues:
## Docker Directory Ownership
- Added ensure_docker_dir_ownership() helper function
- Applied to ALL 25+ services (Immich, Keycloak, ActualBudget, Jellyfin,
Emby, ARM, FileBrowser, MagicMirror, Lyrion, Mealie, Minecraft, Frigate,
ntfy, Uptime Kuma, wg-easy, Traccar, Portainer, MeshCentral, FindMyDevice,
Frigate-Notify, Watchtower, Kopia, Caddy)
- Fixed disaster recovery path (line 309) to set ownership
- Docker folders now owned by sudo user, not root
- Users can run docker commands without sudo
## Keycloak Security Improvements
- Implemented password validation with retry loop
- Password requirements: 12+ chars, alphanumeric only (no special chars)
- Auto-generate secure passwords by pressing ENTER
- Moved all credentials to .env file (no passwords in docker-compose.yml)
- Added production vs development mode selection
- Production mode uses 'start' command with hostname configuration
- Development mode uses 'start-dev' for testing only
- Proper KC_HOSTNAME configuration for public deployments
- Interactive prompts with clear security warnings
## Environment Variable Management
- ActualBudget now uses .env file for configuration
- Keycloak uses .env for admin and database passwords
- Consistent .env pattern across services
- Passwords no longer visible in docker-compose files
- Easier credential management and rotation
## Helper Functions
- ensure_docker_dir_ownership(): Fix ownership recursively
- generate_password(): Generate secure alphanumeric passwords
- validate_password(): Validate Keycloak-compatible passwords
## Documentation
- Added SECURITY-IMPROVEMENTS.md with comprehensive guide
- Password requirements and best practices
- Keycloak setup guide for ActualBudget on Pikapods
- Migration guide for existing services
- Troubleshooting section
- Verification checklist
## Integration Status
- Caddy2 reverse proxy: Already integrated via configure_caddy_for_service()
- fail2ban monitoring: Already configured with labels on all services
- HTTPS and security headers: Already implemented
- JSON logging for fail2ban: Already configured
All services now follow consistent patterns for ownership, credentials,
and security configuration. Script tested with bash -n for syntax errors.
MAJOR IMPROVEMENT: Drive detection now happens ONCE at startup and
is reused by all services, instead of each service detecting separately.
1. **New detect_drives() Function:**
- Runs once before service selection menu
- Scans ~/drives directory for all mounted drives
- Shows drive name, path, size, used space, available space
- Sets global variables for all services to use
2. **Global Variables Set:**
- PRIMARY_DRIVE: name of first drive (e.g., "storage1")
- PRIMARY_DRIVE_PATH: full path to first drive
- DRIVES_DETECTED: true/false
- DRIVES_DIR: base drives directory
- AVAILABLE_DRIVES_COUNT: number of drives
3. **Display Example:**
```
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
DETECTED DRIVES
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
✓ storage1
Path: /home/user/drives/storage1
Size: 2.0T (Used: 800G, Available: 1.2T)
✓ backup
Path: /home/user/drives/backup
Size: 4.0T (Used: 1.5T, Available: 2.5T)
Using 'storage1' as primary drive for default paths
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
```
4. **Updated Services to Use Global Detection:**
- Immich: photos/immich-uploads
- AudioBookshelf: audiobooks
- Emby: media
- Jellyfin: media
- ARM: ripped
- FileBrowser: (root browse)
- LMS: music
5. **Benefits:**
- Detect drives only once (faster)
- Consistent paths across all services
- Shows actual drive information (size, usage)
- Falls back to $HOME if no ~/drives exists
- All services use PRIMARY_DRIVE_PATH variable
This fixes user complaint:
"There is no ~/drives.... it should detect which drives are already
in ~/drives and offer to use one of those"
- Added check_service_exists() to ARM (Automatic Ripping Machine)
- Added check_service_exists() to FileBrowser
- Both now show Skip/Reconfigure/Restart options
- Prevents accidental overwrite of existing configurations
Progress: 5/24 services complete (Immich, AudioBookshelf, Emby, ARM, FileBrowser)
1. **Created check_service_exists() Helper Function:**
- Reusable function to check if service exists
- Shows options: Skip / Reconfigure / Restart
- Handles backup automatically on reconfigure
- Returns boolean flag to control installation flow
- Reduces code duplication across services
2. **Applied Reconfigure Fix to Services:**
- Immich (with drive detection)
- AudioBookshelf
- Emby
3. **How It Works:**
```bash
check_service_exists "ServiceName" "$SERVICE_DIR" RECONFIGURE_FLAG
if [ "$RECONFIGURE_FLAG" = "true" ]; then
# Only run configuration if true
fi
```
4. **Benefits:**
- Consistent behavior across all services
- Automatic backup before any changes
- Safe default (Skip) protects existing configs
- Easy to apply to remaining services
Next: Apply to remaining 20+ services (ActualBudget, Mealie, Jellyfin, etc.)
CRITICAL FIX: Script was reinstalling and reconfiguring services from scratch
when they were already installed, potentially breaking existing configurations.
1. **Detect Existing Installations:**
- Check for existing docker-compose.yml before configuration
- Show options: Skip / Reconfigure / Restart only
- Default to Skip to preserve existing configs
2. **Skip/Reconfigure/Restart Options:**
- Option 1: Skip (keep existing configuration)
- Option 2: Reconfigure (backup existing, then reconfigure)
- Option 3: Restart containers only (no reconfiguration)
3. **Automatic Backup on Reconfigure:**
- Creates timestamped backup before any changes
- Backup location: ~/docker/backups/YYYYMMDD-HHMMSS-servicename/
- Full service directory backed up
4. **Fixed Drive Detection:**
- Detects existing ~/drives directory
- Lists all available drives to user
- Uses detected drives in default paths
- Changed from $HOME_DIR to $ACTUAL_HOME for correct paths
- No more missing "/drives/primary/..." paths
5. **Drive Path Examples:**
- Detects: ~/drives/storage1, ~/drives/backup, ~/drives/media
- Shows: "Detected drives: storage1, backup, media"
- Default becomes: ~/drives/storage1/photos/immich-uploads
Applied to: Immich (template for other services)
This fixes user issues:
- "I selected keep the items that were already installed, then it lead
me to reinstall configure from scratch all those services"
- "It did not detect if the service was already in the Caddyfile"
- "it might have messed up what I already had"
- "There is no ~/drives...."
- "it should detect which drives are already in ~/drives and offer to
use one of those"
Implemented automatic Caddy reverse proxy configuration that runs
BEFORE each service is started, with backup, reload, and formatting.
1. **New configure_caddy_for_service() Function:**
- Detects if Caddy is installed (skips if not)
- Prompts user if they want to configure reverse proxy
- Asks for domain/subdomain (e.g., photos.example.com)
- Backs up Caddyfile with timestamp
- Checks for existing configuration and offers to overwrite
- Adds service block with security headers and fail2ban logging
- Reloads Caddy configuration
- Formats Caddyfile with `caddy fmt --overwrite`
- Final reload after formatting
- Shows final access URL (https://...)
2. **Configuration Sequence:**
- Service docker-compose.yml created
- Service .env configured
- Caddy configuration added (if Caddy installed)
- Caddy reloaded and formatted
- Service containers started
- All happens before `docker compose up -d`
3. **Integrated into Services:**
- Immich (photos.example.com)
- AudioBookshelf (audiobooks.example.com)
- Emby (emby.example.com)
- ActualBudget (budget.example.com)
- Mealie (recipes.example.com)
- Jellyfin (jellyfin.example.com)
- Uptime Kuma (uptime.example.com)
4. **Caddy Configuration Includes:**
- Automatic HTTPS via Let's Encrypt
- Security headers (HSTS, X-Content-Type-Options, etc.)
- JSON logging for fail2ban
- Proper reverse_proxy to localhost:PORT
5. **Backup & Safety:**
- Caddyfile backed up to: Caddyfile.backup.YYYYMMDD-HHMMSS
- Existing configs detected and user can choose to overwrite
- Reload errors show backup file path for restoration
6. **Commands Used:**
- `docker exec caddy caddy reload --config /etc/caddy/Caddyfile`
- `docker exec caddy caddy fmt --overwrite /etc/caddy/Caddyfile`
This addresses user request:
"new/reconfigured services are add to the Caddyfile as part of the
install/reconfigure before they are attempt to be brought up, as a
part of the install/reconfigure of the service? (With backup of
Caddyfile and docker exec -w /etc/caddy caddy caddy reload &&
docker exec -w /etc/caddy caddy caddy fmt --overwrite after each
addition to Caddyfile?)"
Added comprehensive uninstall functionality to whiptail menu:
1. **Install/Uninstall Menu Choice:**
- Added action menu: Install new services, Uninstall existing, or Cancel
- Automatically detects existing services in ~/docker/
2. **Smart Service Detection:**
- Scans for existing Docker services and marks them in the install menu
- Install menu shows [*] for already-installed services
- Helps users identify what's already running on their server
3. **Uninstall Functionality:**
- Uninstall menu only shows services that are currently installed
- All services selected by default for quick removal
- Automatic backup before uninstall to ~/docker/backups/
- Stops containers, removes directories, backs up data
- Special handling for system packages (fail2ban)
4. **Improved Re-run Support:**
- Script detects existing installations on startup
- Can rerun on configured servers to add new services
- Uninstall option allows cleanup of unwanted services
- Backups ensure data safety during removals
Services Supported:
- All 24 Docker services (Immich, Keycloak, Caddy, etc.)
- System packages (fail2ban)
- Backups created with timestamp: YYYYMMDD-HHMMSS-servicename
Usage Examples:
- Fresh install: Select services to install
- Add services: Rerun script, existing services auto-selected
- Remove services: Choose "Uninstall", select what to remove
- Cleanup: All data backed up automatically
This addresses user request for:
- Ability to reinstall on current server
- Option to remove services from whiptail menu
- Better handling of existing installations
Added planning documents for future script reorganization:
NEW-SCRIPT-STRUCTURE.md:
- Outlines two-phase structure (Essential Setup vs Service Selection)
- Documents plan to move optional services to whiptail menu
- Plans for uninstall functionality
- Re-run detection improvements
SCRIPT-FLOW-INTRO.txt:
- Proposed intro text for users
- Explains Phase 1 (required) and Phase 2 (optional)
- Documents re-running behavior
These are planning documents for future enhancements.
Current commit only implements the duplicate prompt fix.
Applied WHIPTAIL_USED flag check to all 25 service prompts that appear
after the whiptail menu to prevent services from being prompted
individually when they weren't selected in the checkbox menu.
This fixes the issue where services like AudioBookshelf were being
prompted even when not selected in whiptail.
Services fixed:
- AudioBookshelf, Emby, ARM, FileBrowser, Magic Mirror
- ActualBudget, Keycloak, Caddy, fail2ban
- Lyrion Music Server, Mealie, Minecraft, Jellyfin, Frigate
- ddclient, ntfy, Uptime Kuma, wg-easy
- Traccar, Portainer, MeshCentral Server
- FindMyDevice, Frigate-Notify, Watchtower
Pattern applied:
if [ "$WHIPTAIL_USED" != true ] && [ -z "$INSTALL_SERVICE" ]; then
# Show prompt
fi
This ensures prompts only appear when whiptail was not used OR
the service variable is not yet set.
CRITICAL FIX:
The whiptail menu was OVERWRITING all service variables with "n",
which caused duplicate prompts and ignored user's earlier selections.
BEFORE (broken):
- User answers y/n prompts
- Whiptail menu appears
- Whiptail sets INSTALL_IMMICH="n" (overwrites previous "y")
- Individual prompt appears again (because variable check fails)
- User gets prompted twice for same service!
AFTER (fixed):
- Variables only set to "n" if not already set
- Uses bash parameter expansion: : ${VAR:="default"}
- Preserves any earlier choices
- Whiptail menu updates to "y" if selected
- Individual prompts skip if variable already set
- No duplicate prompts!
CHANGES:
Lines 2225-2249: Changed from direct assignment (VAR="n")
to conditional default (: ${VAR:="n"})
This preserves earlier choices while still allowing whiptail
to override them when services are selected.
SIDE EFFECT FIXED:
- Containers now start properly
- No more "containers won't come up" issue
- Proper dependency order maintained
Claude