I have a Qnap (yes, I hear the boo’s) that I’m trying to figure out an easy way to backup my linux box to. All the solutions I find are CLI and I’d rather have a GUI to do it so I’m not fucking everything up.
I have a good nas to cloud app, but I need to get this thing on the nas first.
Happy to take suggestions on where to look. I’ve been googling for about an hour and haven’t found what I need. Either it’s just the OS or direct form the desktop to the cloud. Any ideas?
Take a look at https://freefilesync.org/.
The ones I usually rely on (SyncBackPro, SyncTime, GoodSync, etc.) have no Linux version, or are CLI only. But FreeFileSync is easy to install on Rocky and has a decent GUI.
Awesome! Thanks Jan! That worked a treat!
hate to be that guy,. but once you learn about the power of rsync youll soon preffer the cli for stuff like this
bascially
rsync -avhp /source/you/want/to/copy/ /target/you/copy/to/
I love it deeply personally.
Essentially what I’m looking for is getting as close as possible to a full disk clone. Will rsync copy everything on the drive? I have the nas loading on boot from the fstab so it shouldn’t be an issue, except when the copy comes to the /mnt folder. I had to filter out /mnt and /proc.
What are the -avhp tags representing?
Others can weigh in here…
But in my mind there are three data sets to worry about, and what your goal is.
What I’m not certain about is, whether you could boot of a different drive and rsync back your NAS to your boot drive and re-establish a working system. Also keeping in mind that rsync does not keep any history, but is always just the latest, possibly too recent to recover from a failure.
For full system recovery, an actual disk snapshot seems like a more foolproof way of doing that, rather than a filesystem level rsync.
That said, and rsync of your user’s home folder, plus any data drives you have mounted, maybe even the /opt file tree with all the apps is totally usable. In that case you quickly re-install a virgin OS and rsync all the non-OS data back in, and it should fit.
That leaves a gap with all the little changes we make to the OS (tweaks, mounts, etc.). While they may not rsync back cleanly, having a copy of files like /etc/fstab for reference so you don’t do it from scratch is super valuable. As snapshots are in-time, you could keep a few of thema round for different check points.
Mac TimeMachine is great for restoring a borked OS. But that’s because it’s tightly integrated with the recovery function in the boot loader. I’m not sure if there’s a good Rocky equivalent?
PS: Was looking more into that. I’ve come across a few articles that describe full system backup/recovery with rsync. However, it think they assume the OS lives on a single disk. My Rocky 8.7 seems seems to be split over 3 filesystems (root, home, and boot). So you would have to rsync them separately, or your system will look differently afterwards. You also have to restore the LVM config on the drives.
So yes, rsync might be able to do full system recovery on a straightforward single, non-LVM disk config.
Explainshell has this awesome tool that allows you to type in a command with flags and it’ll tell you all the information you need to know about these flags.
https://explainshell.com/explain?cmd=rsync+-avhp
If you were a Synology user, I would tell you to use Synology Active Business Backup. A little Google fu tells me that Qnaps equivalent is Netbak Replicator.
Something with snapshots is ideal. They capture moments in time, each hour, then, each day, then, each week, then each month. While our Sink is super fast and dope, and I highly recommend that you just know that tool in general, you’d have to upgrade to rsnapshot in order to get snapshots of rsyncs.
This is a great way to ensure against user error, hardware failures, and malware.
Thanks Jan and Randy. I looked in to Netbak but they only have a windows option. I also really like explainshell. Takes the mystery out of things.
I think what I want is to have a full image of my new system with everything installed and all the tweaks and fidgets done to make it a workable machine before I start mucking with things. I’ll look in to finding something that does rsync copies on a daily basis. I snapshot my nas every night, but I want to make sure that I’m keeping the computer backups up to date as well.
For true, just role it back kind of backups, there a tools that come on bootable memory sticks that can clone a drive at the physical level. A bit slower, but good of you want to preserve your system in time.
Another option is Ansible. It’s an open-source and easy way to run repeatable commands to perform tasks. Once you setup an SSH key between an Ansible controller (I use a Rocky VM in VirtualBox) and your machine(s), and set a super simple Ansible hosts file that specifies groups of machines, you can create something like this…
install_step1_dku.yml
install_step2_chrome_fstab_teradici_mounts_tailscale_plugins.yml
install_step3_teradici.yml
install_step4_tailscale.yml
install_step5_flame_2023.3.yml
install_step5_flame_2024.yml
install_step6_user_mltw.yml
install_step7_zfs.yml
These are Ansible Playbook files, that I created steps…
This is what an Ansible Playbook file looks like… it magically installs the DKU
---
- name: Install DKU
hosts: all
become: yes
gather_facts: no
tasks:
- name: Install NFS utils
ansible.builtin.yum:
name: nfs-utils
state: present
become: yes
- name: Create /Volumes/hotstorage directory
ansible.builtin.file:
path: /Volumes/hotstorage
state: directory
become: yes
- name: Mount NFS share
ansible.builtin.mount:
path: /Volumes/hotstorage
src: '10.10.0.122:volume1/hotstorage'
fstype: nfs
opts: 'defaults'
state: mounted
become: yes
- name: Copy DKU tarball
ansible.builtin.copy:
src: "/Volumes/hotstorage/02_assets/flame_versions/DKU_18.1.0.tar"
dest: "/tmp/DKU_18.1.0.tar"
remote_src: yes
- name: Extract DKU tarball
ansible.builtin.unarchive:
src: "/tmp/DKU_18.1.0.tar"
dest: "/tmp"
remote_src: yes
- name: Install DKU
ansible.builtin.shell: "/tmp/DKU_18.1.0/INSTALL_DKU"
args:
chdir: "/tmp/DKU_18.1.0"
register: install_dku_result
- name: Reboot the system
ansible.builtin.reboot:
This one installs all the plugins I care about…
- name: Install Chrome and update /etc/fstab
hosts: linux
become: true
tasks:
- name: Install dnf-plugins-core
ansible.builtin.package:
name: dnf-plugins-core
state: present
- name: Install epel-release
ansible.builtin.package:
name: epel-release
state: present
- name: Enable powertools repository
ansible.builtin.command: dnf config-manager --set-enabled powertools
- name: Install wget and dnf-utils
ansible.builtin.package:
name:
- wget
- dnf-utils
state: present
- name: Download Google Chrome RPM
ansible.builtin.get_url:
url: https://dl.google.com/linux/direct/google-chrome-stable_current_x86_64.rpm
dest: /tmp/google-chrome-stable_current_x86_64.rpm
- name: Install Google Chrome
ansible.builtin.dnf:
name: /tmp/google-chrome-stable_current_x86_64.rpm
state: present
disable_gpg_check: yes
- name: Add NFS entry to /etc/fstab
ansible.builtin.lineinfile:
path: /etc/fstab
line: '10.10.0.122:/volume1/hotstorage /Volumes/hotstorage nfs soft,timeo=60,_netdev,rw 0 0'
- name: Ensure /Volumes/hotstorage directory exists
ansible.builtin.file:
path: /Volumes/hotstorage
state: directory
- name: Mount all filesystems
ansible.builtin.mount:
fstype: nfs
opts: soft,timeo=60,_netdev,rw
src: 10.10.0.122:/volume1/hotstorage
path: /Volumes/hotstorage
state: mounted
- name: Add Tailscale repository
ansible.builtin.command: dnf config-manager --add-repo https://pkgs.tailscale.com/stable/centos/8/tailscale.repo
- name: Install Tailscale
ansible.builtin.package:
name: tailscale
state: present
- name: Enable and start tailscaled
ansible.builtin.systemd:
name: tailscaled
enabled: yes
state: started
# - name: Bring Tailscale up
# ansible.builtin.shell: tailscale up
# register: tailscale_up_output
# changed_when: false
# - name: Display Tailscale up output
# ansible.builtin.debug:
# msg: "{{ tailscale_up_output.stdout }}"
# - name: Get Tailscale IP
# ansible.builtin.shell: tailscale ip -4
# register: tailscale_ip_output
# changed_when: false
# - name: Display Tailscale IP
# ansible.builtin.debug:
# msg: "{{ tailscale_ip_output.stdout }}"
# - name: Install NeatVideo5OFX
# ansible.builtin.shell:
# cmd: "/Volumes/hotstorage/02_assets/installers/NeatVideo5OFX.Pro.Intel64/NeatVideo5OFX.Pro.Intel64.run"
- name: Install SapphireOFX package
ansible.builtin.yum:
name: /Volumes/hotstorage/02_assets/installers/SapphireOFX-2022.500-1.0.x86_64.rpm
state: present
disable_gpg_check: yes
- name: Install MochaPro2022 package
ansible.builtin.yum:
name: /Volumes/hotstorage/02_assets/installers/MochaPro2022-9.0.1-49.g18e5e8f59b54.x86_64.rpm
state: present
disable_gpg_check: yes
- name: Install MochaPro2022-OFX package
ansible.builtin.yum:
name: /Volumes/hotstorage/02_assets/installers/MochaPro2022-OFX-9.0.1-49.g18e5e8f59b54.rhel.rpm
state: present
disable_gpg_check: yes
- name: Check if Neat Video is already installed
stat:
path: "/usr/local/Neat Video v5 OFX"
register: neat_video_installed
- name: Run Neat Video installer if not installed
ansible.builtin.shell: "/Volumes/hotstorage/02_assets/installers/NeatVideo5OFX.Pro.Intel64/NeatVideo5OFX.Pro.Intel64.run --mode silent"
args:
chdir: "/Volumes/hotstorage/02_assets/installers/NeatVideo5OFX.Pro.Intel64"
when: not neat_video_installed.stat.exists
- name: Enable chrony service
systemd:
name: chronyd
enabled: yes
state: started
- name: Set timezone to automatic
shell: timedatectl set-ntp true
It looks complicated but I did this entirely using Chat GPT. I kinda gave up on Linux backups because I was always tweaking and figured if the shit hits the fan, all I want is a quick way of getting up and running with repeatable steps that I could automate during a freakout. And for me, in a weird way, redoing something automatically is somehow almost as god as having a backup of it in the first place. I dunno. I’m weird like that.
I know it’s not what you were asking for, but, it might be worth thinking about long term. I mean, heck, I’ll give you these and you could change the paths.
Another rad ansible source…
