Skip to main content

Practical DevOps scripts

Table of contents​

  1. Automated deployment
  2. Backup and restore
  3. Monitoring and alerts
  4. Log management
  5. System maintenance
  6. CI/CD scripts

1 - Automated deployment​

Simple deployment script​

#!/bin/bash
set -euo pipefail

#######################
# Configuration
#######################
readonly APP_NAME="myapp"
readonly DEPLOY_DIR="/var/www/$APP_NAME"
readonly BACKUP_DIR="/var/backups/$APP_NAME"
readonly GIT_REPO="[email protected]:user/myapp.git"
readonly BRANCH="${1:-main}"

#######################
# Fonctions
#######################

log() {
    echo "[$(date '+%Y-%m-%d %H:%M:%S')] $*"
}

backup_current() {
    if [ -d "$DEPLOY_DIR" ]; then
        local backup_name="${APP_NAME}_$(date '+%Y%m%d_%H%M%S')"
        log "Backup vers $BACKUP_DIR/$backup_name"
        cp -r "$DEPLOY_DIR" "$BACKUP_DIR/$backup_name"
    fi
}

deploy() {
    log "Déploiement de la branche: $BRANCH"
    
    # Clone ou pull
    if [ -d "$DEPLOY_DIR/.git" ]; then
        cd "$DEPLOY_DIR"
        git fetch origin
        git checkout "$BRANCH"
        git pull origin "$BRANCH"
    else
        git clone -b "$BRANCH" "$GIT_REPO" "$DEPLOY_DIR"
        cd "$DEPLOY_DIR"
    fi
    
    # Install dependencies
    if [ -f "package.json" ]; then
        log "Installation des dépendances npm"
        npm ci --production
    fi
    
    if [ -f "requirements.txt" ]; then
        log "Installation des dépendances Python"
        pip install -r requirements.txt
    fi
}

restart_service() {
    log "Redémarrage du service"
    sudo systemctl restart "$APP_NAME" || true
}

healthcheck() {
    log "Vérification de santé"
    local max_attempts=30
    local attempt=1
    
    while [ $attempt -le $max_attempts ]; do
        if curl -sf http://localhost:8080/health > /dev/null; then
            log "Application OK"
            return 0
        fi
        log "Tentative $attempt/$max_attempts..."
        sleep 2
        ((attempt++))
    done
    
    log "ERREUR: L'application ne répond pas"
    return 1
}

#######################
# Main
#######################

main() {
    log "=== Début du déploiement ==="
    
    backup_current
    deploy
    restart_service
    
    if ! healthcheck; then
        log "ROLLBACK nécessaire"
        exit 1
    fi
    
    log "=== Déploiement terminé ==="
}

main

Zero-downtime deployment​

#!/bin/bash
set -euo pipefail

# Blue-Green deployment simplifié
readonly APP_DIR="/var/www"
readonly CURRENT_LINK="$APP_DIR/current"
readonly RELEASES_DIR="$APP_DIR/releases"

deploy_zero_downtime() {
    local release_dir="$RELEASES_DIR/$(date '+%Y%m%d%H%M%S')"
    
    # Préparer la nouvelle release
    mkdir -p "$release_dir"
    git clone "$GIT_REPO" "$release_dir"
    cd "$release_dir"
    npm ci --production
    npm run build
    
    # Basculer le lien symbolique (atomique)
    ln -sfn "$release_dir" "$CURRENT_LINK"
    
    # Recharger l'application
    sudo systemctl reload nginx
    
    # Nettoyer les anciennes releases (garder les 5 dernières)
    ls -1dt "$RELEASES_DIR"/* | tail -n +6 | xargs -r rm -rf
}

🔝 Back to table of contents​


2 - Backup and restore​

Complete backup script​

#!/bin/bash
set -euo pipefail

#######################
# Configuration
#######################
readonly BACKUP_BASE="/backup"
readonly DATE=$(date '+%Y%m%d')
readonly RETENTION_DAYS=30

# Ce qu'on sauvegarde
declare -A BACKUP_SOURCES=(
    [web]="/var/www"
    [config]="/etc"
    [data]="/home"
)

# Base de données
readonly DB_NAME="myapp"
readonly DB_USER="backup"
readonly DB_PASS_FILE="/root/.db_password"

#######################
# Fonctions
#######################

log() {
    echo "[$(date '+%Y-%m-%d %H:%M:%S')] $*" | tee -a "$BACKUP_BASE/backup.log"
}

backup_files() {
    local name="$1"
    local source="$2"
    local dest="$BACKUP_BASE/$DATE/${name}.tar.gz"
    
    log "Backup fichiers: $name"
    
    mkdir -p "$(dirname "$dest")"
    tar -czf "$dest" -C "$(dirname "$source")" "$(basename "$source")" 2>/dev/null || {
        log "WARN: Certains fichiers ignorés dans $name"
    }
    
    log "  -> $(du -h "$dest" | cut -f1)"
}

backup_database() {
    local dest="$BACKUP_BASE/$DATE/database.sql.gz"
    
    log "Backup base de données: $DB_NAME"
    
    PGPASSWORD=$(cat "$DB_PASS_FILE") pg_dump \
        -U "$DB_USER" \
        -h localhost \
        "$DB_NAME" | gzip > "$dest"
    
    log "  -> $(du -h "$dest" | cut -f1)"
}

cleanup_old_backups() {
    log "Nettoyage des backups de plus de $RETENTION_DAYS jours"
    
    find "$BACKUP_BASE" -maxdepth 1 -type d -mtime +$RETENTION_DAYS -exec rm -rf {} \;
}

verify_backup() {
    local dir="$BACKUP_BASE/$DATE"
    
    log "Vérification des backups"
    
    for file in "$dir"/*.tar.gz; do
        if tar -tzf "$file" > /dev/null 2>&1; then
            log "  OK: $(basename "$file")"
        else
            log "  ERREUR: $(basename "$file")"
            return 1
        fi
    done
    
    if gunzip -t "$dir/database.sql.gz" 2>/dev/null; then
        log "  OK: database.sql.gz"
    else
        log "  ERREUR: database.sql.gz"
        return 1
    fi
}

#######################
# Main
#######################

main() {
    log "=== Début du backup ==="
    
    # Backup des fichiers
    for name in "${!BACKUP_SOURCES[@]}"; do
        backup_files "$name" "${BACKUP_SOURCES[$name]}"
    done
    
    # Backup de la base
    backup_database
    
    # Vérification
    verify_backup
    
    # Nettoyage
    cleanup_old_backups
    
    # Résumé
    log "=== Backup terminé ==="
    log "Taille totale: $(du -sh "$BACKUP_BASE/$DATE" | cut -f1)"
}

main

Restore script​

#!/bin/bash
set -euo pipefail

restore_backup() {
    local backup_date="${1:?Usage: $0 <date YYYYMMDD>}"
    local backup_dir="/backup/$backup_date"
    
    [ -d "$backup_dir" ] || {
        echo "Backup non trouvé: $backup_dir"
        exit 1
    }
    
    echo "Restauration du backup: $backup_date"
    read -p "Confirmer? (y/N) " -n 1 -r
    echo
    [[ $REPLY =~ ^[Yy]$ ]] || exit 0
    
    # Restaurer les fichiers
    for archive in "$backup_dir"/*.tar.gz; do
        echo "Restauration: $(basename "$archive")"
        tar -xzf "$archive" -C /
    done
    
    # Restaurer la base
    echo "Restauration base de données"
    gunzip -c "$backup_dir/database.sql.gz" | psql -U postgres myapp
    
    echo "Restauration terminée"
}

restore_backup "$@"

🔝 Back to table of contents​


3 - Monitoring and alerts​

Monitoring script​

#!/bin/bash
set -uo pipefail

#######################
# Configuration
#######################
readonly SLACK_WEBHOOK="${SLACK_WEBHOOK:-}"
readonly EMAIL_TO="${ALERT_EMAIL:[email protected]}"

# Seuils
readonly DISK_THRESHOLD=80
readonly MEMORY_THRESHOLD=90
readonly LOAD_THRESHOLD=4

#######################
# Fonctions
#######################

send_alert() {
    local level="$1"
    local message="$2"
    
    echo "[$level] $message"
    
    # Slack
    if [ -n "$SLACK_WEBHOOK" ]; then
        curl -sf -X POST -H 'Content-type: application/json' \
            -d "{\"text\":\"[$level] $(hostname): $message\"}" \
            "$SLACK_WEBHOOK" > /dev/null || true
    fi
    
    # Email
    echo "$message" | mail -s "[$level] Alert from $(hostname)" "$EMAIL_TO" 2>/dev/null || true
}

check_disk() {
    local usage
    usage=$(df / | awk 'NR==2 {print int($5)}')
    
    if [ "$usage" -gt "$DISK_THRESHOLD" ]; then
        send_alert "WARNING" "Disque: ${usage}% utilisé (seuil: ${DISK_THRESHOLD}%)"
        return 1
    fi
    return 0
}

check_memory() {
    local usage
    usage=$(free | awk '/Mem:/ {printf "%.0f", $3/$2 * 100}')
    
    if [ "$usage" -gt "$MEMORY_THRESHOLD" ]; then
        send_alert "WARNING" "Mémoire: ${usage}% utilisée (seuil: ${MEMORY_THRESHOLD}%)"
        return 1
    fi
    return 0
}

check_load() {
    local load
    load=$(cat /proc/loadavg | cut -d' ' -f1)
    
    if (( $(echo "$load > $LOAD_THRESHOLD" | bc -l) )); then
        send_alert "WARNING" "Load: $load (seuil: $LOAD_THRESHOLD)"
        return 1
    fi
    return 0
}

check_service() {
    local service="$1"
    
    if ! systemctl is-active --quiet "$service"; then
        send_alert "CRITICAL" "Service $service est DOWN!"
        return 1
    fi
    return 0
}

check_url() {
    local url="$1"
    local name="${2:-$url}"
    
    if ! curl -sf --connect-timeout 10 "$url" > /dev/null; then
        send_alert "CRITICAL" "$name ne répond pas!"
        return 1
    fi
    return 0
}

#######################
# Main
#######################

main() {
    local status=0
    
    check_disk || status=1
    check_memory || status=1
    check_load || status=1
    
    # Services
    for service in nginx postgresql; do
        check_service "$service" || status=1
    done
    
    # URLs
    check_url "http://localhost/health" "App Health" || status=1
    
    exit $status
}

main

🔝 Back to table of contents​


4 - Log management​

Rotation and analysis​

#!/bin/bash
set -euo pipefail

# Rotation manuelle des logs
rotate_logs() {
    local log_dir="${1:-/var/log/myapp}"
    local max_files="${2:-7}"
    
    for log in "$log_dir"/*.log; do
        [ -f "$log" ] || continue
        
        # Rotation
        for i in $(seq $((max_files-1)) -1 1); do
            [ -f "${log}.$i" ] && mv "${log}.$i" "${log}.$((i+1))"
        done
        
        [ -f "$log" ] && mv "$log" "${log}.1"
        touch "$log"
    done
    
    # Compression des anciens
    find "$log_dir" -name "*.log.[2-9]" -exec gzip {} \;
}

# Analyse des erreurs
analyze_errors() {
    local log_file="$1"
    local period="${2:-1 hour ago}"
    
    echo "=== Erreurs depuis $period ==="
    
    awk -v since="$(date -d "$period" '+%Y-%m-%d %H:%M')" '
        $1 " " $2 >= since && /ERROR|CRITICAL|FATAL/ {
            count[$0]++
        }
        END {
            for (err in count) {
                printf "%5d: %s\n", count[err], err
            }
        }
    ' "$log_file" | sort -rn | head -20
}

# Top des IPs
top_ips() {
    local access_log="${1:-/var/log/nginx/access.log}"
    local n="${2:-10}"
    
    echo "=== Top $n IPs ==="
    awk '{print $1}' "$access_log" | sort | uniq -c | sort -rn | head -n "$n"
}

🔝 Back to table of contents​


5 - System maintenance​

Maintenance script​

#!/bin/bash
set -euo pipefail

log() {
    echo "[$(date '+%H:%M:%S')] $*"
}

# Mise à jour système
update_system() {
    log "Mise à jour des packages"
    apt-get update -qq
    apt-get upgrade -y -qq
    apt-get autoremove -y -qq
}

# Nettoyage Docker
cleanup_docker() {
    if command -v docker &> /dev/null; then
        log "Nettoyage Docker"
        docker system prune -af --volumes || true
    fi
}

# Nettoyage fichiers temporaires
cleanup_temp() {
    log "Nettoyage fichiers temporaires"
    find /tmp -type f -atime +7 -delete 2>/dev/null || true
    find /var/tmp -type f -atime +30 -delete 2>/dev/null || true
}

# Nettoyage logs
cleanup_logs() {
    log "Nettoyage logs anciens"
    find /var/log -name "*.gz" -mtime +30 -delete 2>/dev/null || true
    journalctl --vacuum-time=7d || true
}

# Vérification disque
check_disk_health() {
    log "Vérification santé disque"
    for disk in /dev/sd[a-z]; do
        [ -e "$disk" ] || continue
        smartctl -H "$disk" 2>/dev/null || true
    done
}

# Main
main() {
    log "=== Maintenance système ==="
    
    update_system
    cleanup_docker
    cleanup_temp
    cleanup_logs
    check_disk_health
    
    log "=== Maintenance terminée ==="
}

main

🔝 Back to table of contents​


6 - CI/CD scripts​

Build script​

#!/bin/bash
set -euo pipefail

readonly BUILD_DIR="build"
readonly VERSION="${CI_COMMIT_TAG:-dev-$(git rev-parse --short HEAD)}"

build() {
    echo "Building version: $VERSION"
    
    rm -rf "$BUILD_DIR"
    mkdir -p "$BUILD_DIR"
    
    # Build application
    npm ci
    npm run build
    
    # Copy to build dir
    cp -r dist/* "$BUILD_DIR/"
    
    # Create version file
    echo "$VERSION" > "$BUILD_DIR/VERSION"
}

test_app() {
    echo "Running tests..."
    npm test
}

package() {
    echo "Creating package..."
    tar -czf "app-${VERSION}.tar.gz" -C "$BUILD_DIR" .
}

main() {
    build
    test_app
    package
    echo "Build complete: app-${VERSION}.tar.gz"
}

main

Test script​

#!/bin/bash
set -euo pipefail

run_tests() {
    echo "=== Running Tests ==="
    
    # Unit tests
    echo "Unit tests..."
    npm run test:unit
    
    # Integration tests
    echo "Integration tests..."
    npm run test:integration
    
    # Lint
    echo "Linting..."
    npm run lint
    
    # Security audit
    echo "Security audit..."
    npm audit --audit-level=high
}

main() {
    run_tests
    echo "All tests passed!"
}

main

🔝 Back to table of contents​


Key takeaways​

  • Deployment scripts with backup and healthcheck
  • Backups with rotation and verification
  • Monitoring with alerts (Slack, email)
  • Log management with analysis and rotation
  • Regular automated maintenance
  • CI/CD scripts for build, test, package

🔝 Back to table of contents​

← Previous chapter | Next chapter: Exercises and projects →