How to Install Linux Without a USB Drive

Installing a new operating system, particularly a robust and versatile one like Linux, traditionally conjures images of bootable USB drives. However, the necessity of a physical USB stick is not a universal requirement. For various reasons, from a lack of available USB ports to a desire for a streamlined installation process, alternative methods exist. This guide explores how to install Linux without relying on a USB drive, focusing on network-based installations and utilizing existing partitions or other media. This approach is particularly relevant for system administrators managing multiple machines, users with older hardware lacking readily accessible USB ports, or those who simply prefer a cleaner, disc-free setup.

Network Installation Methods

The most common and powerful way to install Linux without a USB drive is through network-based installation methods. These leverage the network to deliver the installation files to the target machine. This approach requires a properly configured network and often another machine on the network to act as a server.

PXE Booting (Preboot Execution Environment)

PXE is a standardized booting method that allows a computer to boot from a network interface card (NIC) before loading an operating system from any local storage device. For a PXE installation, you will need:

• A DHCP Server: This server assigns IP addresses to client machines on the network. It must be configured to provide the PXE boot server’s IP address and the boot filename.
• A TFTP (Trivial File Transfer Protocol) Server: TFTP is used to transfer the initial bootloader and kernel files from the PXE server to the client machine.
• An NFS (Network File System) or HTTP Server: Once the initial bootloader has loaded, it needs to access the rest of the installation files. This is typically achieved by mounting a shared directory (via NFS) or downloading files (via HTTP) from a server on the network.
• A PXE-enabled Network Card: Most modern network cards support PXE booting. Ensure it’s enabled in the target machine’s BIOS/UEFI settings.

The Process:

1. Server Setup: Configure your chosen server (often a Linux machine) with a DHCP server (like isc-dhcp-server), a TFTP server (like tftpd-h5), and an NFS or HTTP server. The TFTP server will host the PXE bootloader (e.g., GRUB or ISOLINUX) and the kernel. The NFS/HTTP server will host the entire Linux distribution’s installation ISO contents or a minimal set of files required for network installation.
2. Client Configuration: In the target machine’s BIOS/UEFI, configure the boot order to prioritize network boot. Ensure the NIC is enabled for PXE.
3. Booting the Client: When the target machine boots, it will request an IP address from the DHCP server. The DHCP server will respond with an IP address and the location of the PXE boot server and boot file.
4. Bootloader Download: The client’s NIC will download the bootloader via TFTP.
5. Kernel and Initrd Load: The bootloader will then download the Linux kernel and the initial ramdisk (initrd) from the TFTP server.
6. Installation Begins: The kernel boots, and the initrd environment initializes the network stack. It then connects to the NFS or HTTP server to access the rest of the installation media and begins the standard Linux installation process.

This method is highly scalable and efficient for deploying Linux across many machines simultaneously, making it a cornerstone of enterprise IT and research environments.

Using a Network Install ISO

Some Linux distributions offer specialized “netinstall” or “minimal install” ISO images. These ISOs are significantly smaller than full DVD or Live ISOs because they contain only the bootloader, kernel, and a minimal set of tools needed to initiate a network-based installation.

The Process:

1. Download the Netinstall ISO: Obtain the netinstall ISO image from your chosen Linux distribution’s official website.
2. Burn to CD/DVD (if applicable): If the target machine has an optical drive, you can burn this ISO to a CD or DVD. This bypasses the need for a USB drive entirely.
3. Configure BIOS/UEFI: Set the target machine’s BIOS/UEFI to boot from the CD/DVD drive.
4. Boot and Install: Upon booting from the CD/DVD, the system will load the minimal environment. It will then prompt you to connect to the network. Once a network connection is established (wired is generally more reliable for installations), the installer will download the necessary packages from the distribution’s online repositories and proceed with the installation.

This method is simpler than setting up a full PXE server but still requires the physical media (CD/DVD) or a way to boot from it. It’s an excellent option for individual installations where a USB drive is unavailable but an optical drive is present.

Installation via Existing Partitions or Other Media

Beyond network installations, there are scenarios where you can leverage existing storage or alternative boot methods to install Linux.

Installing from another Linux Partition (Chroot Method)

This method is a bit more advanced and involves installing Linux into a new partition from an already running Linux system. It’s often referred to as a “chroot” installation.

Prerequisites:

• A Running Linux System: This can be a live environment or a pre-existing Linux installation on the target machine or another machine you can access.
• Target Partition: You’ll need an unformatted or formatted partition on the target machine where you intend to install the new Linux system.
• Access to Installation Files: You’ll need the contents of a Linux installation ISO, typically mounted on the running system.

The Process:

1. Mount the Target Partition: Boot into your existing Linux system (or live environment). Identify and mount the target partition where you want to install Linux. For example, if your target partition is /dev/sda3, you would mount it: sudo mount /dev/sda3 /mnt.
2. Mount the Installation Media: Mount the Linux installation ISO on the running system. If you’ve downloaded the ISO, you can mount it like this: sudo mount -o loop /path/to/linux.iso /mnt/iso.
3. Copy Installation Files: Copy the entire contents of the mounted ISO to the target partition. This can be a time-consuming process. sudo cp -a /mnt/iso/* /mnt/.
4. Prepare for Chroot: Before chrooting, you need to mount essential virtual filesystems into the target environment.
   • sudo mount --bind /dev /mnt/dev
   • sudo mount --bind /proc /mnt/proc
   • sudo mount --bind /sys /mnt/sys
5. Chroot into the New Environment: Now, change the root directory to the new installation: sudo chroot /mnt.
6. Configure and Install: Within the chroot environment, you can now run the installation commands. This often involves running a specific installer script provided by the distribution, configuring package managers, setting up users, and installing a bootloader. The exact commands will vary significantly depending on the distribution. For Debian/Ubuntu-based systems, you might use debootstrap or specific installer scripts. For Arch Linux, you would follow its wiki installation guide.
7. Install Bootloader: Crucially, you need to install a bootloader (like GRUB) from within the chroot environment to the target disk’s MBR or EFI partition so that the system can boot into the newly installed Linux.
8. Exit and Reboot: Once the installation and bootloader configuration are complete, exit the chroot environment, unmount the partitions, and reboot the machine.

This method offers granular control but requires a deeper understanding of Linux system administration.

Using a Pre-made Disk Image

For rapid deployment or to replicate a specific Linux setup, you can use a pre-made disk image. This involves creating an image of a fully configured Linux system and then “writing” that image to the target drive.

The Process:

1. Create a Master Image: Install and configure Linux on a “master” machine as desired. Once satisfied, create a disk image of this installation. Tools like dd, Clonezilla, or rsync can be used for this. For example, using dd: sudo dd if=/dev/sda of=/path/to/your/image.img bs=4M status=progress.
2. Transfer the Image: Copy the generated image file to a location accessible by the target machine. This could be a network share, an external hard drive, or another partition.
3. Apply the Image to the Target: Boot the target machine from a live environment or another operating system. Then, write the image file back to the target drive. Using dd again: sudo dd if=/path/to/your/image.img of=/dev/sda bs=4M status=progress. Be extremely careful to ensure you are writing to the correct target drive to avoid data loss.
4. Post-Installation Configuration: After applying the image, you may need to perform some post-installation tasks, such as reconfiguring network interfaces (if MAC addresses are embedded in the image), updating the system, or adjusting hardware-specific settings.

This method is exceptionally fast for deploying identical systems but lacks the flexibility of a standard installation if each machine needs unique configurations.

Conclusion

While the ubiquitous USB drive has become synonymous with operating system installations, it is far from the only method. Network-based installations, particularly PXE booting, offer powerful and scalable solutions for enterprise environments. For individual users, utilizing netinstall ISOs burned to CD/DVDs, or more advanced techniques like chroot installations and disk imaging, provide viable alternatives. Understanding these methods expands your toolkit for deploying and managing Linux systems, offering flexibility and efficiency when traditional USB installations are not an option. Each method has its own set of requirements and complexities, but by mastering them, you can navigate a wider range of installation scenarios with confidence.