In IT, we throw around a lot of acronyms and buzz words! It’s getting to the point now that it’s not uncommon to hear 10 acronyms or made up words in a single sentence. FML… See even I do it!! Object Storage hmmmm
S3, Swift, Ceph, GlusterFS… Do you need it? Maybe, maybe not but a great many people I speak to don’t really know what it is. It can be embarrassing to admit you don’t know the latest buzz word so many bluff their way through. I thought it would be beneficial to go through what it is and then whether you need it.
What is Object Storage
Block and File storage have been around for a while and are fairly well understood.
At its simplest File storage manages and houses the data through a file hierarchy. Block storage manages data as blocks within sectors and tracks. Hence the names Block and File storage.
So, it’s no surprise that Object storage is an architecture that manages data as individual objects. Thus, removing the need for an overarching file system.
What is an Object
What an object is can vary slightly depending on the use case. Typically, it will consist of;
- The data
- Some metadata (structural, descriptive or administrative)
- A global unique identifier
Implementations of Object Storage
Some of the first real use cases for object storage were for archive storage. These systems weren’t after performance just the data services that object storage could offer. Think of systems such as EMC Centerra which was recently End of Life’d. These were eventually replaced by systems we use today such as EMC Atmos and Openstack Swift.
Object based file systems then started to creep into the market. Systems such as Ceph and GlusterFS started to become more commonplace in enterprises. We also see it used as the platform many cloud providers operate in conjunction with Openstack.
Almost all the cloud storage you will see available will be utilising object storage. This is, in my opinion the largest use case. It is used by the likes of Amazon and their S3 service and Zettagrid and its VAST offering.
Actual use cases for businesses
From the people I speak with on a daily basis these are the most common real world use cases. Furthermore, these will likely be what is most noteworthy and especially relevant.
- Backup Software. Integration within your backup software and allow the software to backup natively to an object storage location.
- File Server Replacement. Used in conjunction with 3rd party software to mount as a shared drive to all workstations and remove the need for a shared file server.
- Quorum Disk. Remove the need for a cluster quorum disk for services such as SQL which rely on a witness disk shared across all nodes.
How do I setup Object Storage?
There’s loads of different options out there and you can use what you want. I will show you how to do it on Zettagrid VAST as that’s what I know.
Deploy the service from the Zettagrid portal.
Load up the MyAccount portal and open the newly provisioned service.
Head over to the Security tab and then note down the;
- Access Key
- Secret Key
There you have it, object storage setup in under 5 minutes. Hence, your next step will be driven by how you wish to consume it. As listed above you could enter it directly into your supported backup product. You could run any application that can utilise the RESTful API.
Software that can mount it as a drive such as CloudBerry and others are a great start. For more software that are SWIFT or S3 compatible head here – https://www.zettagrid.com/faqs/applications-vast-compatible/
There we go, you can head into your next IT meeting full bottle on object storage. You will be able to articulate what it is, how to use it and consequently whether your business could benefit from it.
Does your object storage need to be in the same country? Head over here to find out – www.whatwouldlukedo.com/data-sovereignty-data-locality/
I’m keen to hear your thoughts on object storage and how your organisation uses it.