Unlike in the home environment where data can be stored on a single hard drive, organizations have needs to store much larger amounts of data. How much data? We’re talking about really large amounts in the terabytes, petabytes, exabytes and even zettabytes. Also realize that we not only have to store the data primarily, but also have redundancy with multiple copies to ensure that data is not lost. Not only does an organization need to meet its storage needs for today, but it also needs to plan for the future with a scalable solution that grows as the organization does.
Given these challenges, organizations clearly have requirements that extend beyond having a primary hard drive in the computing device, or even a single additional drive via a direct attached storage solution. Rather, they need network storage solutions to handle a colossal amount of drives. Thankfully, there are solutions that can keep up with the needs, and are scalable as well. This brings us to JBOD and RAID, which are the two options in this area.
Depending on who you ask, JBOD stands for either “Just a bunch of disks,” or “Just a bunch of drives.” Semantics aside, JBOD is an expansion unit, as it is an enclosure designed to hold multiple hard drives, and have them connected for use. They get used in data centers, and are a simple to manage way to conveniently add high-density data storage. This is because as more storage is needed, additional drives can be added to the enclosure unit, and when space runs out, the company can either swap out an existing drive for a larger capacity one, which are typically available at a lower cost as time marches on, or purchase the next enclosure and continue to add additional drives.
There are some advantages to a JBOD model of data storage. One is that it is cost effective, as the enclosure can house multiple drives, which can then be added as needed. This also means that it is scalable, as the drives get added over time, up to the capacity of the housing unit. JBOD is also simple to manage, without the need for complex software, and frequent updates. Furthermore, they offer a large amount of storage in a smaller footprint of space. The drives are also hot-swappable, which means that drives can be added or removed without turning off the device, which avoids downtime. A final advantage is the JBOD integrates well with servers, with high performance, and it works well with the gamut of OS’.
A distinguishing point of JBOD is how the data gets stored. It is performed via a process known as spanning. The data as it is received at the enclosure gets written to an individual disk. As this disk fills to capacity, then the next disk receives the data. This allows the enclosure to maximize the capacity to store the largest amount of data. However, the downside is that the data only gets written once, and is vulnerable to failure.
JBOD works well in a number of common use case scenarios. These include as a replacement to tape drive backups, for security and surveillance applications, for data archiving, and for media storage.
The counterpart to JBOD is known as RAID, which stands for a Redundant Array of Independent Disks. Unlike in JBOD where the drives just store the data, in RAID multiple disk drives get combined into a cohesive storage unit. What gets more complicated is that there are several versions of RAID, depending on what the goal is.
The first version is known as RAID 0, which combines at least two drives so that the data is written to both in a process known as ‘Data striping.’ However, between the two drives, there is only a single copy of the data, so that if either drive fails all the data is lost. The goal here is the fastest performance.
A more popular option is called RAID 1, which uses the process of ‘Disk mirroring.’ Here the data gets written simultaneously to at least two drives, and is considered ‘Fault tolerant’ as if either drive fails, there is another copy of the data. In this version, data integrity is favored over pure performance.
The next common version is RAID 5, which requires at least three disks. It essentially combines RAID 0 and RAID 1, to both increase performance and provide drive parity, which allows the data to be reconstructed if any one of the drives suffers a failure. RAID 6 takes this even a step further with a second parity block, which can sustain two disk drive failures, making this even more secure.
As we can see, RAID offers considerably more flexibility and options in comparison to JBOD. This can be achieved through hardware RAID, which has the requirement of a controller within the motherboard slot to connect the drives. A less expensive option is for a software RAID setup, which manages the RAID through a utility application. In general, the hardware RAID offers better performance, but at a higher cost.
RAID offers some advantages over JBOD. This starts with the flexibility, and the option to increase performance, or data redundancy, or both depending on the RAID version chosen. They also allow for hot swapping, allowing a drive to be added, or a failed drive to be replaced, and for the drive to be integrated in. This is particularly powerful as it allows a failed drive to be swapped out, and then backup data written to it, with no downtime.
While RAID costs more than JBOD, and requires more skill and expertise to manage, it is easy to see why it remains a popular option for backing up data.
JBOD and RAID both continue to be viable choices for the storage needs of business. Each has its purposes, balancing performance, and redundancy of storage needs, so consider the advantages in choosing the solution needed for your business.
Are you a pro? Subscribe to our newsletter
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Jonas P. DeMuro is a freelance reviewer covering wireless networking hardware.