308 lines
12 KiB
Bash
308 lines
12 KiB
Bash
#!/bin/bash
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ----------------------------- Weekly Sync Maintenance ----------------------------------------
|
|
# -----------------------------------------------------------------------------------------------
|
|
# Weekly maintenance window orchestrator — clean sync, container updates, weekly restarts.
|
|
# Schedule: 30 2 * * 0 (Sunday 2:30am — fits before 3am network reboot)
|
|
#
|
|
# Execution order:
|
|
# 1. Stop local containers — Emby + auth stack stopped locally
|
|
# 2. Stop remote containers — Emby + auth stack stopped remotely via SSH
|
|
# 3. Pull updates locally — if CRITICAL_SYNC_UPDATES=true
|
|
# 4. Pull updates remotely — if CRITICAL_SYNC_UPDATES_REMOTE=true
|
|
# 5. rsync Emby — full clean mirror, both instances stopped
|
|
# 6. rsync Critical-Data — auth stack clean sync, databases flushed
|
|
# 7. Start remote containers — correct order, delayed start respected
|
|
# 8. Start local containers — correct order, delayed start respected
|
|
# 9. docker_weekly_restart.sh — weekly container restarts
|
|
#
|
|
# Synced shares (WEEKLY_SYNC_JOBS in Master.conf):
|
|
# /mnt/user/Media_Server/Emby — emby profile — full mirror, cache resets weekly
|
|
# /mnt/user/appdata-Failover/Critical-Data — critical-data — auth stack clean state
|
|
#
|
|
# Why weekly instead of nightly for Emby:
|
|
# Emby builds a warm image cache on HOST2 throughout the week
|
|
# Syncing nightly resets cache — cold loads every morning for users
|
|
# Weekly sync: cache stays warm 6 days, resets Sunday night while users sleep
|
|
# emby-failover dirty sync covers watch states + library every 30-60min between syncs
|
|
#
|
|
# Container updates during the window:
|
|
# Containers already stopped for sync — updates pull at zero extra downtime
|
|
# Both servers start on identical image versions after the window completes
|
|
# Toggle: CRITICAL_SYNC_UPDATES / CRITICAL_SYNC_UPDATES_REMOTE in Master.conf
|
|
#
|
|
# What triggers weekly_health_digest.sh:
|
|
# NOT this script — weekly_health_digest.sh runs on its own Saturday schedule
|
|
#
|
|
# Configuration in Master.conf:
|
|
# WEEKLY_SYNC_JOBS — shares synced during the maintenance window
|
|
# WEEKLY_MAINTENANCE_SCRIPTS — scripts run after sync (docker_weekly_restart)
|
|
# CRITICAL_SYNC_UPDATES — toggle container updates on/off
|
|
# CRITICAL_SYNC_UPDATES_REMOTE — toggle remote container updates on/off
|
|
# -----------------------------------------------------------------------------------------------
|
|
# All configuration in Master.conf.
|
|
# Supports --dry-run to walk through without stopping containers, syncing, or updating.
|
|
# -----------------------------------------------------------------------------------------------
|
|
|
|
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
|
|
|
|
source "$SCRIPT_DIR/../Master.conf"
|
|
source "$SCRIPT_DIR/../common.sh"
|
|
|
|
RSYNC_SCRIPT="$SCRIPT_DIR/../Rsync/rsync.sh"
|
|
|
|
parse_args "$@"
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_GEAR Setup ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
echo ""
|
|
echo "━━━ $ICON_GEAR Setup ━━━"
|
|
|
|
if [[ "$EUID" -ne 0 ]]; then
|
|
error "Must be run as root"
|
|
exit 1
|
|
fi
|
|
|
|
success "Running as root"
|
|
|
|
acquire_lock
|
|
|
|
detect_hosts
|
|
resolve_remote_ip
|
|
|
|
# Load container list from emby profile — used for update pulls
|
|
read -r -a MAINTENANCE_CONTAINERS <<< "${PROFILE_CRITICAL_CONTAINER_NAMES[emby]:-} ${PROFILE_CRITICAL_CONTAINER_NAMES[critical-data]:-}"
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_SHIELD Pre-flight Checks ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
echo ""
|
|
echo "━━━ $ICON_SHIELD Pre-flight Checks ━━━"
|
|
|
|
check_connectivity
|
|
check_remote_rootfs
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_STOP $ICON_CONTAINERS Stop Containers ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
echo ""
|
|
echo "━━━ $ICON_STOP $ICON_CONTAINERS Stop Containers ━━━"
|
|
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
warn "DRY RUN — containers will not be stopped"
|
|
else
|
|
# Load critical-data + emby container list for stops
|
|
read -r -a CRITICAL_CONTAINER_NAMES <<< "${PROFILE_CRITICAL_CONTAINER_NAMES[critical-data]:-} ${PROFILE_CRITICAL_CONTAINER_NAMES[emby]:-}"
|
|
read -r -a DELAYED_CONTAINERS <<< "${PROFILE_DELAYED_CONTAINERS[critical-data]:-}"
|
|
CONTAINER_DELAY="${PROFILE_CONTAINER_DELAY[critical-data]:-15}"
|
|
|
|
stop_local_containers
|
|
stop_containers
|
|
fi
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_GEAR Container Updates ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
echo ""
|
|
echo "━━━ $ICON_GEAR Container Updates ━━━"
|
|
|
|
if [[ "$CRITICAL_SYNC_UPDATES" == true ]]; then
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
warn "DRY RUN — would pull updates for local containers"
|
|
for c in "${MAINTENANCE_CONTAINERS[@]}"; do
|
|
[[ -z "$c" ]] && continue
|
|
warn "DRY RUN — would docker pull: $c"
|
|
done
|
|
else
|
|
info "Pulling local container updates..."
|
|
for c in "${MAINTENANCE_CONTAINERS[@]}"; do
|
|
[[ -z "$c" ]] && continue
|
|
# Get image name from running or stopped container
|
|
IMAGE=$(docker inspect "$c" --format '{{.Config.Image}}' 2>/dev/null)
|
|
if [[ -z "$IMAGE" ]]; then
|
|
log "$c — not found locally, skipping update"
|
|
continue
|
|
fi
|
|
info "Pulling $IMAGE for $c..."
|
|
if docker pull "$IMAGE" >/dev/null 2>&1; then
|
|
success "$c — image updated"
|
|
else
|
|
warn "$c — pull failed, will start on existing image"
|
|
fi
|
|
done
|
|
fi
|
|
else
|
|
info "CRITICAL_SYNC_UPDATES=false — skipping local updates"
|
|
fi
|
|
|
|
if [[ "$CRITICAL_SYNC_UPDATES_REMOTE" == true ]]; then
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
warn "DRY RUN — would pull updates on $REMOTE_SERVER_NAME"
|
|
else
|
|
info "Pulling remote container updates on $REMOTE_SERVER_NAME..."
|
|
for c in "${MAINTENANCE_CONTAINERS[@]}"; do
|
|
[[ -z "$c" ]] && continue
|
|
IMAGE=$(ssh -i "$SSH_KEY" root@"$REMOTE_SERVER" \
|
|
"docker inspect $c --format '{{.Config.Image}}' 2>/dev/null" 2>/dev/null)
|
|
if [[ -z "$IMAGE" ]]; then
|
|
log "$c — not found on remote, skipping update"
|
|
continue
|
|
fi
|
|
info "Pulling $IMAGE for $c on $REMOTE_SERVER_NAME..."
|
|
if ssh -i "$SSH_KEY" root@"$REMOTE_SERVER" "docker pull $IMAGE" >/dev/null 2>&1; then
|
|
success "$c — remote image updated"
|
|
else
|
|
warn "$c — remote pull failed, will start on existing image"
|
|
fi
|
|
done
|
|
fi
|
|
else
|
|
info "CRITICAL_SYNC_UPDATES_REMOTE=false — skipping remote updates"
|
|
fi
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_SYNC Critical Shares Sync ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
PASS=()
|
|
FAIL=()
|
|
TOTAL_START=$(date +%s)
|
|
|
|
SYNC_JOBS=("${WEEKLY_SYNC_JOBS[@]}")
|
|
SHARE_COUNT=${#SYNC_JOBS[@]}
|
|
|
|
echo ""
|
|
echo "━━━ $ICON_SYNC Critical Shares Sync — $(date '+%Y-%m-%d %H:%M:%S') ━━━"
|
|
echo "$ICON_SUMMARY Jobs: $SHARE_COUNT"
|
|
echo ""
|
|
|
|
JOB_NUM=0
|
|
for JOB in "${SYNC_JOBS[@]}"; do
|
|
((JOB_NUM++))
|
|
JOB_NAME=$(basename "$JOB")
|
|
echo "━━━ [$JOB_NUM/$SHARE_COUNT] $JOB_NAME ━━━"
|
|
|
|
JOB_START=$(date +%s)
|
|
|
|
# Containers already stopped — rsync profile won't try to stop them again
|
|
# Pass --no-container-stop flag would be ideal but profiles handle this naturally
|
|
# since containers are already stopped, stop_containers finds nothing running
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
bash "$RSYNC_SCRIPT" "$JOB" --dry-run
|
|
else
|
|
bash "$RSYNC_SCRIPT" "$JOB"
|
|
fi
|
|
|
|
EXIT_CODE=$?
|
|
JOB_END=$(date +%s)
|
|
JOB_DURATION=$(format_duration $(( JOB_END - JOB_START )))
|
|
|
|
if [[ "$EXIT_CODE" -eq 0 ]]; then
|
|
PASS+=("$JOB_NAME")
|
|
success "$JOB_NAME — $ICON_SUCCESS done in $JOB_DURATION"
|
|
else
|
|
FAIL+=("$JOB_NAME")
|
|
error "$JOB_NAME — $ICON_ERROR failed after $JOB_DURATION"
|
|
fi
|
|
|
|
echo ""
|
|
done
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_START $ICON_CONTAINERS Start Containers ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
echo ""
|
|
echo "━━━ $ICON_START $ICON_CONTAINERS Start Containers ━━━"
|
|
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
warn "DRY RUN — containers will not be started"
|
|
else
|
|
start_containers
|
|
start_local_containers
|
|
fi
|
|
|
|
TOTAL_END=$(date +%s)
|
|
TOTAL_DURATION=$(format_duration $(( TOTAL_END - TOTAL_START )))
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_GEAR Post-sync Jobs ━━━
|
|
# docker_weekly_restart.sh and any other WEEKLY_MAINTENANCE_SCRIPTS run after sync
|
|
# -----------------------------------------------------------------------------------------------
|
|
JOB_PASS=()
|
|
JOB_FAIL=()
|
|
SCRIPTS_ROOT="$SCRIPT_DIR/.."
|
|
|
|
if [[ ${#WEEKLY_MAINTENANCE_SCRIPTS[@]} -gt 0 ]]; then
|
|
echo ""
|
|
echo "━━━ $ICON_GEAR Post-sync Jobs ━━━"
|
|
|
|
for script_entry in "${WEEKLY_MAINTENANCE_SCRIPTS[@]}"; do
|
|
[[ -z "$script_entry" ]] && continue
|
|
script_args=($script_entry)
|
|
script_path="$SCRIPTS_ROOT/${script_args[0]}"
|
|
script_name=$(basename "${script_args[0]}")
|
|
extra_args=("${script_args[@]:1}")
|
|
|
|
echo ""
|
|
info "$ICON_START Running: $script_name"
|
|
|
|
if [[ ! -f "$script_path" ]]; then
|
|
error "$script_name — not found at $script_path"
|
|
JOB_FAIL+=("$script_name")
|
|
continue
|
|
fi
|
|
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
warn "DRY RUN — would run: $script_name"
|
|
JOB_PASS+=("$script_name (dry run)")
|
|
elif bash "$script_path" "${extra_args[@]}"; then
|
|
success "$script_name — done"
|
|
JOB_PASS+=("$script_name")
|
|
else
|
|
error "$script_name — failed"
|
|
JOB_FAIL+=("$script_name")
|
|
fi
|
|
done
|
|
fi
|
|
|
|
WINDOW_END=$(date +%s)
|
|
|
|
# -----------------------------------------------------------------------------------------------
|
|
# ━━━ $ICON_SUMMARY Summary ━━━
|
|
# -----------------------------------------------------------------------------------------------
|
|
echo ""
|
|
echo "━━━━━ $ICON_SUMMARY WEEKLY SYNC MAINTENANCE SUMMARY ━━━━━"
|
|
echo "$ICON_TIME Duration: $TOTAL_DURATION"
|
|
echo "$ICON_GEAR Updates: local=$CRITICAL_SYNC_UPDATES remote=$CRITICAL_SYNC_UPDATES_REMOTE"
|
|
echo ""
|
|
|
|
echo "$ICON_SYNC Sync jobs:"
|
|
if [[ ${#PASS[@]} -gt 0 ]]; then
|
|
for job in "${PASS[@]}"; do echo " $ICON_SUCCESS $job"; done
|
|
fi
|
|
if [[ ${#FAIL[@]} -gt 0 ]]; then
|
|
for job in "${FAIL[@]}"; do echo " $ICON_ERROR $job"; done
|
|
fi
|
|
echo " Passed: ${#PASS[@]} Failed: ${#FAIL[@]}"
|
|
|
|
if [[ ${#JOB_PASS[@]} -gt 0 || ${#JOB_FAIL[@]} -gt 0 ]]; then
|
|
echo ""
|
|
echo "$ICON_GEAR Post-sync jobs:"
|
|
for job in "${JOB_PASS[@]}"; do echo " $ICON_SUCCESS $job"; done
|
|
for job in "${JOB_FAIL[@]}"; do echo " $ICON_ERROR $job"; done
|
|
fi
|
|
|
|
echo ""
|
|
|
|
TOTAL_FAIL=$(( ${#FAIL[@]} + ${#JOB_FAIL[@]} ))
|
|
|
|
if [[ "$DRY_RUN" == true ]]; then
|
|
echo "$ICON_WARN Status: DRY RUN — no changes made"
|
|
elif [[ "$TOTAL_FAIL" -eq 0 ]]; then
|
|
echo "$ICON_DONE Status: $ICON_SUCCESS ALL COMPLETE"
|
|
notify "Weekly sync maintenance complete on $(hostname) — synced + updated (local=$CRITICAL_SYNC_UPDATES remote=$CRITICAL_SYNC_UPDATES_REMOTE)" "Weekly Maintenance" "normal"
|
|
else
|
|
echo "$ICON_ERROR Status: $TOTAL_FAIL failure(s) — check logs"
|
|
notify "Weekly sync maintenance failed on $(hostname) — sync: ${#FAIL[@]} failed, jobs: ${#JOB_FAIL[@]} failed" "Weekly Maintenance" "warning"
|
|
fi
|
|
echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━" |