"🗄️ Complete Server Clone & Backup — Linux"
If you like this project, consider supporting me on Buy Me a Coffee ☕️
tags:
- backups
- s3
- sysadmin
-
linux
Server cloning via tar + SSH is a classic sysadmin technique for migrating or duplicating Linux servers by streaming a compressed filesystem archive directly over the network. It preserves permissions, ownership, and special files. This approach is commonly used for quick server migrations, disaster recovery, and bare-metal cloning. Modern alternatives include rsync (incremental), Clonezilla (disk-level), and Ansible/Terraform (infrastructure-as-code). / Клонирование сервера через tar + SSH — классический метод миграции или дублирования Linux-серверов путём потоковой передачи сжатого архива файловой системы по сети. Современные альтернативы: rsync, Clonezilla, Ansible/Terraform.
Installation & Configuration
Dependencies
# Debian/Ubuntu
apt update && apt install tar openssh-client # Install tar + SSH / Установка tar + SSH
# RHEL/CentOS/AlmaLinux
yum install tar openssh-clients # Install tar + SSH / Установка tar + SSH
SSH Performance Tuning
# Lighter cipher for fast networks (less secure — use only in trusted LAN)
#
ssh -oCiphers=aes128-ctr <USER>@<HOST>
# Disable SSH compression (faster on fast networks)
ssh -oCompression=no <USER>@<HOST>
Core Management (Tar & Backups)
Basic Backup Commands
# Full system backup excluding virtual/system directories
#
tar -cvpzf "backup-$(date +%m-%d-%Y-%H%M).tar.gz" \
--exclude=/backup.tar.gz \
--exclude=/dev --exclude=/boot --exclude=/proc --exclude=/sys \
--exclude=/mnt --exclude=/lost+found \
--one-file-system /
# Backup specific directory (e.g., web root)
tar -czvf "backup-$(date +%m-%d-%Y-%H%M).tar.gz" \
--exclude={./public_html/templates/cache,./public_html/templates/compiled,./public_html/images} \
./public_html
Unpacking & Restoring
tar -xvf backup.tar.gz # Extract archive / Распаковать архив
tar -C /dest -xvf backup.tar.gz # Extract to specific directory / В конкретную папку
Server Cloning (Piped Transfer)
[!IMPORTANT] This method clones the filesystem directly across the network over SSH. Ensure the target server has identical or compatible hardware (especially for the bootloader). Run from the source server only.
Clone Entire Server via SSH
[!CAUTION] This overwrites everything on the target server. Always verify
<REMOTE_IP>before running. There is no undo.
# Execute on SOURCE server
cd / && \
tar cvpzf - \
--exclude=/dev \
--exclude=/boot \
--exclude=/etc/fstab \
--exclude=/etc/mtab \
--exclude=/etc/blkid \
--exclude=/etc/modprobe.conf \
--exclude=/proc \
--exclude=/mnt \
--exclude=/sys \
--exclude=/lost+found \
--exclude=/etc/sysconfig/network-scripts/* \
--ignore-failed-read \
/ | ssh -T root@<REMOTE_IP> 'tar -C / -xvpz'
Cleanup After Cloning
# Remove temporary maildrop files
find /var/spool/postfix/maildrop/ -type f -exec rm {} \;
Database Management (MySQL/MariaDB)
Secure Credential Management
[!TIP] Never use
-p<PASSWORD>directly in shell commands or scripts — it appears in process listings and shell history. Usemysql_config_editoror.my.cnfinstead.
Using mysql_config_editor (Recommended
# Create login profile (password prompted interactively)
mysql_config_editor set --login-path=backup --host=localhost --user=root --password
# Use login path for dumps
mysqldump --login-path=backup --all-databases | gzip > all-databases.sql.gz
Using .my.cnf
/root/.my.cnf
[client]
user=<USER>
password=<PASSWORD>
chmod 600 /root/.my.cnf # Restrict permissions / Ограничить права
Backup & Restore
# Dump all databases
mysqldump -u root -p --all-databases | gzip > "/mysql-backup-$(date +%F).sql.gz"
# Restore from SQL file
mysql -u root -p < file.sql
# Restore from compressed SQL
gunzip -cd mysql-backup.sql.gz | mysql -u root -p
Automation & Scripts
Integrated Backup Script
/usr/local/bin/server-backup.sh
#!/bin/bash
# Automated backup script: DB dump + filesystem archive
#
set -euo pipefail
BACKUP_DIR="/backup"
mkdir -p "$BACKUP_DIR"
rm -f "$BACKUP_DIR"/*.zip "$BACKUP_DIR"/*.tar.gz 2>/dev/null || true
NOW=$(date +%d%m%Y-%H%M%S)
FILENAME="server-backup"
FULL_PATH="$BACKUP_DIR/$FILENAME-$NOW"
# Database dump
# NOTE: Use --login-path for production
mysqldump --no-tablespaces -y -u <USER> -p<PASSWORD> -h <SOURCE_IP> <DB_NAME> > "$FULL_PATH.sql"
# Compress SQL dump
zip -j "$FULL_PATH.zip" "$FULL_PATH.sql"
rm "$FULL_PATH.sql"
# Create system archive
tar -cvpzf "$FULL_PATH.tar.gz" \
--exclude=/var/spool/postfix/maildrop/ \
--exclude=/dev --exclude=/boot \
--exclude=/proc --exclude=/sys \
--exclude=/mnt --exclude=/lost+found \
--exclude=/swap \
--exclude=/var/lib/mysql/ \
--exclude=/etc/sysconfig/network-scripts/* \
--exclude="$BACKUP_DIR"/* \
--one-file-system /
# Optional: Upload to remote server
# scp -P 22 "$FULL_PATH.tar.gz" backup_user@<HOST>:/path/to/backups/
echo "$(date): Backup complete: $FULL_PATH"
chmod +x /usr/local/bin/server-backup.sh
Cron Schedule
/etc/cron.d/server-backup
# Full server backup daily at 01:00
0 1 * * * root /usr/local/bin/server-backup.sh >> /var/log/server-backup.log 2>&1
Logrotate / Logrotate
/etc/logrotate.d/server-backup
/var/log/server-backup.log {
daily
rotate 30
compress
delaycompress
missingok
notifempty
create 640 root root
}
Advanced Operations (Multi-volume)
Using tarcat
Used to concatenate GNU tar multi-volume archives / Объединение многотомных архивов GNU tar.
/usr/local/bin/tarcat
#!/bin/sh
# Usage: tarcat volume1 volume2 ... | tar -xf -
# Author: Bruno Haible, Sergey Poznyakoff
dump_type() {
dd if="$1" skip=${2:-0} bs=512 count=1 2>/dev/null | tr '\0' ' ' | cut -c157
}
case $(dump_type "$1") in
[gx]) PAX=1;;
esac
cat "$1"
shift
for f; do
SKIP=0
T=$(dump_type "$f")
if [ -n "$PAX" ]; then
if [ "$T" = "g" ]; then
SKIP=5
fi
else
if [ "$T" = "V" ]; then T=$(dump_type "$f" 1); fi
if [ "$T" = "M" ]; then SKIP=$((SKIP + 1)); fi
fi
dd skip=$SKIP if="$f"
done
chmod +x /usr/local/bin/tarcat
# Usage
tarcat archive-* | tar -xf -
Comparison of Cloning Approaches
| Method | Pros / Плюсы | Cons / Минусы |
|---|---|---|
| Tar over SSH | Fast, simple, preserves permissions exactly | Manual cleanup of system files (fstab, network). No deduplication |
| Rsync | Incremental sync, resume capability, bandwidth efficient | Slightly slower initial sync than tar |
| Restic / Borg | Deduplication, encryption, snapshot history | Requires client installation, more complex setup |
| Clonezilla / DD | Bit-perfect clone including bootloader | Usually requires downtime; large image if not sparse |
Sysadmin Operations
Network & IP Updates After Cloning
# Bulk replace old IP in config files
find /etc/ -type f -exec sed -i 's/<SOURCE_IP>/<REMOTE_IP>/g' {} \;
Path & Log Reference
/var/log/syslog # System log (Debian/Ubuntu) / Системный лог
/var/log/messages # System log (RHEL) / Системный лог
/var/log/cron # Cron job log / Лог задач cron
/var/log/server-backup.log # Script log (if configured) / Лог скрипта
Default Ports
# SSH: 22
# MySQL/MariaDB: 3306
Service Management
systemctl restart sshd # Restart SSH / Перезапустить SSH
systemctl status mysql # Check MySQL status / Проверить MySQL
[!CAUTION] Always verify your backups are restorable! Run a test restore to a staging server regularly. A backup is only as good as its last verified restoration.
Documentation
- rsync Documentation: https://man7.org/linux/man-pages/man1/rsync.1.html