To some, object storage systems have been something off in the distance. Other people may use it, but they’re not sure it really helps them. Add the perceived requirement of changing user behavior and you have a stalemate. But what if object storage can truly solve one or more problems in your data center? Is that possible? Let’s take a look.
There are two problems that object storage systems can easily solve: the difficulty of finding unstructured data in a sea of unstructured files, and the limits of a budget and the accompanying never ending list of requirements.
First let’s discuss the problem of unstructured data sprawl by looking at two common types of unstructured data: user-generated data and automatically-created data. User-generated data grows at an ever increasing rate and no one wants to delete anything. The problem with this is twofold: it wastes space on a primary storage system and the backup system that protects it, and files created more than a few days ago are quite difficult to find. If a person who created a given file cannot remember where they put it, chances are it will be lost forever. This unfortunately means that a significant portion of primary and backup storage is taken up with files that noone will ever use again.
The second type of unstructured data is automatically-created data such as video surveillance, sensor data, and seismic information. Metadata is the key to storing and finding this data which makes it easy to find using searches such as “show me all of the video surveillance from these cameras on these dates.” All of this data simply needs a safe place for it to store and allow its retrieval using metadata. Typical filesystems can satisfy the first requirement but rarely can they satisfy the second one; additional software will be required to search this automatically created data.
This brings us to the second problem, which is that all of this data that needs storage on a system that does not easily facilitate its retrieval is also on a very expensive system. But it is easy to justify the costs of the typical primary storage and the data protection system when the data the company is creating is a tool to generate revenue. However, when the storage system seems to do a whole lot of storing and not a lot of retrieving, one can begin to question whether its cost is justifiable.
If any part of the problems mentioned above ring a bell, then an object storage system can certainly help. The first reason is that an object storage system by design stores and retrieves objects by its content and metadata. It is easy to search for objects with common metadata, which makes retrieving data at any time much easier than with traditional filesystems. In addition, object storage systems are much less expensive than traditional storage devices. Add to that the fact that object storage systems can protect themselves without the use of a backup system, and you have even more savings.
Object storage systems can certainly help customers easily store and retrieve unstructured data without many of the issues that we find in typical filesystem devices. If those storing user-generated files in such a system are able to more easily find their previous work and reuse it, significant savings are possible. The cost savings of object storage over traditional devices would also allow for the storage of more automatically created data that has other value to the business. Since this type of data does not typically add immediately to the bottom line, the cost of storing it should be kept to a minimum, and an object storage system easily accomplishes that. If these problems resonate then the answer to the title question is yes, an object storage system can help.
Sponsored by Dell EMC®
Dell EMC®’s Elastic Cloud Storage (ECS™), provides a cloud-scale object-based platform that delivers rich metadata search and management capabilities. Users can leverage ECS’s metadata to tag files with important details and easily search through data to provide the quickest time to results.