4.0 2008-04-01
Introduction to Block Devices Partitions

Although it is theoretically possible to use a full disk to house your Linux system, this is almost never done in practice. Instead, full disk block devices are split up in smaller, more manageable block devices. On systems, these are called partitions.

Itanium systems use EFI, the Extensible Firmware Interface, for booting. The partition table format that EFI understands is called GPT, or GUID Partition Table. The partitioning program that understands GPT is called "parted", so that is the tool we will use below. Additionally, EFI can only read FAT filesystems, so that is the format to use for the EFI boot partition, where the kernel will be installed by "elilo".

Advanced Storage

The Installation CDs provide support for EVMS and LVM2. EVMS and LVM2 increase the flexibility offered by your partitioning setup. During the installation instructions, we will focus on "regular" partitions, but it is still good to know EVMS and LVM2 are supported as well.

Designing a Partitioning Scheme Default Partitioning Scheme

If you are not interested in drawing up a partitioning scheme for your system, you can use the partitioning scheme we use throughout this book:

/dev/sda1vfat32MEFI Boot partition/dev/sda2(swap)512MSwap partition/dev/sda3ext3Rest of the diskRoot partition
Partition Filesystem Size Description

If you are interested in knowing how big a partition should be, or even how many partitions you need, read on. Otherwise continue now with partitioning your disk by reading Using parted to Partition your Disk.

How Many and How Big?

The number of partitions is highly dependent on your environment. For instance, if you have lots of users, you will most likely want to have your /home separate as it increases security and makes backups easier. If you are installing Gentoo to perform as a mailserver, your /var should be separate as all mails are stored inside /var. A good choice of filesystem will then maximise your performance. Gameservers will have a separate /opt as most gaming servers are installed there. The reason is similar for /home: security and backups. You will definitely want to keep /usr big: not only will it contain the majority of applications, the Portage tree alone takes around 500 Mbyte excluding the various sources that are stored in it.

As you can see, it very much depends on what you want to achieve. Separate partitions or volumes have the following advantages:

  • You can choose the best performing filesystem for each partition or volume
  • Your entire system cannot run out of free space if one defunct tool is continuously writing files to a partition or volume
  • If necessary, file system checks are reduced in time, as multiple checks can be done in parallel (although this advantage is more with multiple disks than it is with multiple partitions)
  • Security can be enhanced by mounting some partitions or volumes read-only, nosuid (setuid bits are ignored), noexec (executable bits are ignored) etc.

However, multiple partitions have one big disadvantage: if not configured properly, you might result in having a system with lots of free space on one partition and none on another. There is also a 15-partition limit for SCSI and SATA.

As an example partitioning, we show you one for a 20GB disk, used as a demonstration laptop (containing webserver, mailserver, gnome, ...):

$ df -h
Filesystem    Type    Size  Used Avail Use% Mounted on
/dev/sda5     ext3    509M  132M  351M  28% /
/dev/sda2     ext3    5.0G  3.0G  1.8G  63% /home
/dev/sda7     ext3    7.9G  6.2G  1.3G  83% /usr
/dev/sda8     ext3   1011M  483M  477M  51% /opt
/dev/sda9     ext3    2.0G  607M  1.3G  32% /var
/dev/sda1     ext2     51M   17M   31M  36% /boot
/dev/sda6     swap    516M   12M  504M   2% <not mounted>
(Unpartitioned space for future usage: 2 GB)

/usr is rather full (83% used) here, but once all software is installed, /usr doesn't tend to grow that much. Although allocating a few gigabytes of disk space for /var may seem excessive, remember that Portage uses this partition by default for compiling packages. If you want to keep /var at a more reasonable size, such as 1GB, you will need to alter your PORTAGE_TMPDIR variable in /etc/make.conf to point to the partition with enough free space for compiling extremely large packages such as OpenOffice.

Using parted to Partition your Disk

The following parts explain how to create the example partition layout described previously, namely:

/dev/sda1EFI Boot partition/dev/sda2Swap partition/dev/sda3Root partition
Partition Description

Change your partition layout according to your own preference.

Viewing the Current Partition Layout

parted is the GNU partition editor. Fire up parted on your disk (in our example, we use /dev/sda):

# parted /dev/sda

Once in parted, you'll be greeted with a prompt that looks like this:

GNU Parted 1.6.22
Copyright (C) 1998 - 2005 Free Software Foundation, Inc.
This program is free software, covered by the GNU General Public License.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without
even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
General Public License for more details.

Using /dev/sda
(parted)

At this point one of the available commands is help, which you should use if you want to see the other available commands. Another command is print which you should type next to display your disk's current partition configuration:

(parted) print
Disk geometry for /dev/sda: 0.000-34732.890 megabytes
Disk label type: gpt
Minor    Start       End     Filesystem  Name                  Flags
1          0.017    203.938  fat32                             boot
2        203.938   4243.468  linux-swap
3       4243.469  34724.281  ext3

This particular configuration is very similar to the one that we recommend above. Note on the second line that the partition table is type is GPT. If it is different, then the ia64 system will not be able to boot from this disk. For the sake of this guide we'll remove the partitions and create them anew.

Removing all Partitions Unlike fdisk and some other partitioning programs which postpone committing changes until you give the write instruction, parted commands take effect immediately. So once you start adding and removing partitions, you can't simply quit without writing them... they've already been written.

The easy way to remove all partitions and start fresh, which guarantees that we are using the correct partition type, is to make a new partition table using the mklabel command. After you do this, you will have an empty GPT partition table.

(parted) mklabel gpt
(parted) print
Disk geometry for /dev/sda: 0.000-34732.890 megabytes
Disk label type: gpt
Minor    Start       End     Filesystem  Name                  Flags

Now that the partition table is empty, we're ready to create the partitions. We will use a default partitioning scheme as discussed previously. Of course, don't follow these instructions to the letter if you don't want the same partitioning scheme!

Creating the EFI Boot Partition

We first create a small EFI boot partition. This is required to be a FAT filesystem in order for the ia64 firmware to read it. Our example makes this 32 megabytes, which is appropriate for storing kernels and elilo configuration. You can expect each ia64 kernel to be around 5 megabytes, so this configuration leaves you some room to grow and experiment.

(parted) mkpart primary fat32 0 32
(parted) print
Disk geometry for /dev/sda: 0.000-34732.890 megabytes
Disk label type: gpt
Minor    Start       End     Filesystem  Name                  Flags
1          0.017     32.000  fat32
Creating the Swap Partition

Let's now create the swap partition. The classic size to make the swap partition was twice the amount of RAM in the system. In modern systems with lots of RAM, this is no longer necessary. For most desktop systems, a 512 megabyte swap partition is sufficient. For a server, you should consider something larger to reflect the anticipated needs of the server.

(parted) mkpart primary linux-swap 32 544
(parted) print
Disk geometry for /dev/sda: 0.000-34732.890 megabytes
Disk label type: gpt
Minor    Start       End     Filesystem  Name                  Flags
1          0.017     32.000  fat32
2         32.000    544.000
Creating the Root Partition

Finally, let's create the root partition. Our configuration will make the root partition to occupy the rest of the disk. We default to ext3, but you can use ext2, jfs, reiserfs or xfs if you prefer. The actual filesystem is not created in this step, but the partition table contains an indication of what kind of filesystem is stored on each partition, and it's a good idea to make the table match your intentions.

(parted) mkpart primary ext3 544 34732.890
(parted) print
Disk geometry for /dev/sda: 0.000-34732.890 megabytes
Disk label type: gpt
Minor    Start       End     Filesystem  Name                  Flags
1          0.017     32.000  fat32
2         32.000    544.000
3        544.000  34732.874
Exiting parted

To quit from parted, type quit. There's no need to take a separate step to save your partition layout since parted has been saving it all along. As you leave, parted gives you reminder to update your /etc/fstab, which we'll do later in this guide.

(parted) quit
Information: Don't forget to update /etc/fstab, if necessary.

Now that your partitions are created, you can continue with Creating Filesystems.

Creating Filesystems Introduction

Now that your partitions are created, it is time to place a filesystem on them. If you don't care about what filesystem to choose and are happy with what we use as default in this handbook, continue with Applying a Filesystem to a Partition. Otherwise read on to learn about the available filesystems...

Applying a Filesystem to a Partition

To create a filesystem on a partition or volume, there are tools available for each possible filesystem:

vfatmkdosfsext2mke2fsext3mke2fs -jreiserfsmkreiserfsxfsmkfs.xfsjfsmkfs.jfs
Filesystem Creation Command

For instance, to have the boot partition (/dev/sda1 in our example) as vfat and the root partition (/dev/sda3 in our example) as ext3, you would run the following commands:

# mkdosfs /dev/sda1
# mke2fs -j /dev/sda3
Activating the Swap Partition

mkswap is the command that is used to initialize swap partitions:

# mkswap /dev/sda2

To activate the swap partition, use swapon:

# swapon /dev/sda2

Create and activate the swap with the commands mentioned above.

Mounting

Now that your partitions are initialized and are housing a filesystem, it is time to mount those partitions. Use the mount command. Don't forget to create the necessary mount directories for every partition you created. As an example we mount the root and boot partition:

# mount /dev/sda3 /mnt/gentoo
Unlike some of the other architectures supported by Gentoo, /boot is not mounted on ia64. The reason for this is that the EFI boot partition will be automatically mounted and written by the elilo command each time that you run it. Because of this, /boot resides on the root filesystem and is the storage place for the kernels referenced by your elilo configuration. If you want your /tmp to reside on a separate partition, be sure to change its permissions after mounting: chmod 1777 /mnt/gentoo/tmp. This also holds for /var/tmp.

We will also have to mount the proc filesystem (a virtual interface with the kernel) on /proc. But first we will need to place our files on the partitions.

Continue with Installing the Gentoo Installation Files.