Linux Disk Partition

Posted on Apr 11, 2018 in Technology

In Linux, there are a few different partitioning command line tools we can use. One that supports both MBR and GPT partitioning is the parted tool. Parted can be used in two modes. The first is interactive, meaning we’re launched into a separate program, like when we use the less command. The second is command line, meaning you just run commands while still in your shell. We’re going to be using the interactive mode for most of this lesson. Before we do that let’s run a command to show what disks are connected to the computer using the command line mode. We can do this by running the parted – l command. So sudo parted – l. This lists out the disks that are connected to our computer. We can see that the disk /dev/sda is 128 gigabytes. I’ve also plugged in a USB drive and you can see that, /dev /sdb is around 8 gigabytes. Let’s quickly go through what this output says. Here we can see the partition table is listed as gpt. The number field corresponds to the number of partitions on the disk. We can see that there are three partitions. Since this disk is /dev/sta, the first partition will correspond to /dev/sta 1 and the second will correspond to /dev/sta 2 et cetera. The start field is where the partition starts on the disk. For this first partition we can see that it starts at 1,049 kilobytes and ends at 538 megabytes. The field after that shows us how large the partition size is. The next field tells us what file system is on the partition. Then, we have the name and finally, we can see some flags that are associated with this partition. You can see here that /dev /sdb doesn’t currently have any partitions, we’ll fix that in a minute. Let’s select our /dev/sdb disk and start partitioning it. We want to be super careful that we select the correct disk when partitioning something so we don’t accidentally partition the wrong disk. We’re going to use the interactive mode of parted by running sudo parted /dev/sdb. Now we’re in the parted tool. From here, we can run more commands. If we want to get out of this tool and go back to the shell then we just use the quick command. I’m going to run print just to see this disk one more time. It says we have an unrecognized disk label. We’ll need to set a disk label with the mklabel command. Since we want to use the gpt partition table let’s use this command. Mklabel gpt. Let’s look at the status of our disk again to do that we can use a print command. Here we can see the disk information for the selected /dev/sdb disk. Now it says we have the partition table gpt. All right. Let’s start making modifications to the disk. We want to partition the /dev/sdb disk into two partitions. Inside the parted tool we’re going to use the mkpart command. The mkpart command needs to have the following information, what type partition we want to make, what file system we want to format, and the start of the disk and the end of the disk like this.

3:47

The partition type is only meaningful for mbr partition tables. Remember, the mbr uses primary, extended, and logical partitions. Since we are formatting this using gpt, we’re just going to use primary as the partition type. The start point here is one mebibyte and the endpoint is five gibibytes. So our partition is essentially five gibibytes. Remember from the earlier course, that data sizes have long been referred to in two different ways, using the exact data measurement and the estimated data measurement. Remember that one kibibyte is actually 1,024 bytes while one kilobyte is 1,000 bytes. We haven’t really had to care about this distinction before. Some operating systems sometimes measure one kilobyte as 1,024 bytes which is confusing, but when dealing with data storage we want to make sure we’re using the precise measurements so we don’t waste precious storage space. Let’s opt to use mebibyte and gibibyte in our partition. Next, we’re going to format the partition with the file system using mkfs. So I’m just going to quick, sudo mkfs type is ext4. And I want to format the partition, so sdb1. We also left the rest of the disk unpartitioned because we’re going to use it for something else later. With that, we’ve created a partition and formatted a file system on a USB drive. Remember to always be careful when using the parted tool. It’s very powerful and if you modify the wrong disk on here it could cause a pretty big mess. Even though we’ve partitioned our disk and formatted a file system on here, we’re not actually able to start reading and writing files to it just yet. There’s one last step to get a usable disk in Linux. We have to mount the file system to a directory so that we can access it from the shell. Spoiler alert, you’ll learn how to do that in the next video.

In Linux, the dedicated area of the hard drive used for virtual memory is known as swap space. We can create swap space by using the new disk partitioning tools that we learned. A good guideline to use to determine how much swap space you need is to follow the recommended partitioning scheme in the next supplementary reading. In our case, since we just have a USB drive which doesn’t need swap, we’re just going to partition the rest of it as swap to show you how this works. In practice, you would create swap partitions for your main storage devices like hard drives and SSDs. Okay. Let’s make swap space. First, go back into the parted tool and select /dev/sdb, where our USB is. We’re going to partition it again this time to make a swap partition. And then we’ll format the Linux dash swap file system on it. So, mkpart primary Linux swap 5 gibibytes 100 percent. You’ll notice that the end point of the drive says 100 percent which indicates that we should use the rest of the free space on our drive. We’re not done yet. Swap isn’t actually a file system, so this command won’t be enough. I know I’m sorry, I just lied to you like five seconds ago. If you think about it, it makes a lot of sense since pages go into swap and not file data. Anyways, to complete this process, we need to specify that we want to make it swap space with the mkswap command. Let’s quit out of parted and run this command on a new swap partition. So, sudo mkswap dev, and our new swap partition is on dev sdb2. Finally, there’s one more command to run to enable swap on the device, swapon. So, sudo swapon dev sdb2. If we want to automatically mount swap space every time the computer boots up, just add a swap entry to the /etc fstab file like we did earlier.