Using a storage partition in your data storage infrastructure

Learn about the common use cases for a storage partition in a data storage infrastructure, as well as the differences between a storage partition and a LUN.

A partition enables the separation of a hard disk so you can have different operating systems on the same hard disk for file management. But what exactly is a storage partition? In this podcast, Greg Schulz, founder and senior analyst at Stillwater, Minn.-based StorageIO Group, discusses the use cases for a storage partition. He also examines the differences between a drive partition and a logical unit number (LUN), and provides some common examples of storage partitions in a data storage infrastructure.

You can read a transcript of the interview below or listen to the MP3.

Play now:

Download for later:

Using a storage partition
• Internet Explorer: Right Click > Save Target As
• Firefox: Right Click > Save Link As What is a storage partition and what are some use cases for a storage partition?

Schulz: Storage partition has a lot of different meanings; consequently, there's a lot of confusion around [the term]. What it comes down to is that a partition is a way of dividing storage. If you say a drive partition, a disk partition, it's taking a 500 Gb drive and making it into two 250 Gb partitions. If you have a Windows PC, for example, or you have a Windows laptop, you're probably already using drive partitioning and may not realize it. That familiar C: or D: drive, or if there's a backup of your operating system for doing restores, that's an example of a drive-level partition. What is the difference between a drive partition and a LUN?

Schulz: In some ways they're very similar, but I'm sure some would debate they're the same or totally different. What they have in common is that they're both a means of taking storage and carving it up for different purposes -- for example, in that analogy I just gave about a Windows-type system where you take part of the drive for where you're going to put your data, and part of the drive for where you have your backup or recovery. Or if you're on a Unix or a Linux-type system, where you set up different partitions where different file systems get mounted into.

But where things start to differentiate is when you're partitioning, or think of it as subdividing, an individual drive. When you're taking disk drives and putting them into a storage system into a RAID system -- whether it be iSCSI, SAN, NAS, Fibre Channel -- that's where you're taking multiple drives and aggregating them, RAID-protecting and striping them, and then carving up those volumes into logical unit numbers. A LUN is that address by which the storage is actually accessed; the SCSI target or the command for how you get to the storage. So does that mean a drive partition is a storage system and that a virtual machine is a LUN?

Schulz: You have a drive, which could be partitioned, but then a drive could also be mapped into what we'll call a RAID group or volume group. Different vendors call them different things. That grouping could be subdivided as a LUN and presented back to either a physical or virtual machine. In other words, it looks like that operating system might see it as 'Oh, that's the C: drive' or 'That's Target 0 drive.' That operating system, in turn, can take and do different things with it. For example, if you're on Windows, Windows might see several different LUNs, several different targets that it can take, pull in, stripe, map and do different things with.

Where it kind of gets confusing is that you actually have different layers of aggregating, striping, pooling, grouping, RAID protecting but also subdividing that you can actually put layer upon layer upon layer. But there was an interesting piece to your question, which was about a virtual machine. Keep in mind that a virtual machine in a virtual server is a disk file when it's not loaded into memory. That disk file is saved on a storage device, either on a file system, a LUN on that target or on a partition on a drive. Can you give me some examples of storage partitions?

Schulz: Let's break it into a couple of pieces. First, a disk partition, like the example we've already used. In other words, in a Windows-based system, that disk drive that is in that computer, laptop or server that's divided into what you see as the C: or D: drive. This is probably one of the most common examples out there. Or, if you're on a Unix or Linux system, it's taking that drive in that server and dividing that into different pieces where there are different malpoints, different file systems installed.

The other example would be where we've taken some disk drives, put them into a RAID array, into a storage system, created a RAID set -- maybe a RAID 5 or RAID 6 -- striping, mirroring, whatever it happens to be, presenting all the different LUNs, but using a feature that's in some storage systems that has the ability of creating a partition -- you can almost think of it as a virtual storage machine. This is where it can get kind of confusing with virtual servers. What it's doing is that some storage systems have the ability to take those LUNs we talked about out of those groups, take those LUNs, and map or mask them back to different servers. Think of it this way: If I have four servers, and those four servers are all accessing a common, shared storage system, I need a means of isolating and protecting. This is typically done using volume mapping, volume masking, LUN mapping and LUN masking, and also via techniques such as partitioning, where each of those different servers thinks it has its own Target 0, LUN 0.

Dig Deeper on Data storage strategy