Skip to main content

Hi there,

In this article, I’m going to look at why, as the title says, it’s now crucial to give your infrastructure object storage for writing backups.

To do this, though, we need to look at a few things. I’m going to run through them now.

 

Why Enterprise Backups Are Becoming More and More Important

 

In recent years, the amount of data being generated has increased a lot. Companies now need to manage information that is critical to their daily operations. At the same time, with this growth, there are also a variety of risks associated with different situations, such as:

Cybersecurity: Ransomware and data breaches are happening more often and are getting more sophisticated.
Regulations like GDPR and NIS2 require us to keep data safe and secure.
Business Continuity: Losing data can cause a lot of problems for a company. It can damage their reputation, cause financial losses, and disrupt their operations.

 

The answer is on-premises object storage.

On-prem object storage is a pretty cool solution for getting around some of the limitations of traditional backups. For example:

Scalability: It can handle a lot of data without slowing down.
Efficiency: You’ll cut down on management costs and get data to people faster.
Security: You’re in complete control of where and how your data is stored.
Compatibility: It’s easy to integrate with backup software, too.
Immutability: The backup data is immutable throughout the specified retention period.

 

Let’s take a look at the differences between object, block, and file storage.


1. Object Storage

Data are grouped into units called objects, with each object made up of:

Data: is the actual content.
Metadata: detailed info about the data, like the creation date or permissions.
A unique identifier: is an index that allows you to access the data quickly.

Here are the main features:

You can scale it up as much as you need to, which is great for storing huge amounts of unstructured data.
You can usually access the API via S3-compatible APIs that are optimized for distributed access.
This is what most people use it for: backing up data, storing media content, logs, and IoT data.

 

2. Block Storage

The data are organized in blocks of a fixed size. You’ll often hear these referred to as solutions in storage area network (SAN) environments, where the blocks are managed at a low level by the operating system.

Here are the main features:

You can configure each block independently, which is pretty flexible.
It’s great for high-performance tasks like databases or virtual machines.
The operating system manages it, but it doesn’t include specific metadata like object storage.

 

3. File Storage

Your data is organized in a hierarchical file system, which is made up of folders and files. This is the traditional storage model, which is mainly used as NAS (network-attached storage).

Here are the main features:

You can access it via network protocols like SMB, NFS, or FTP.
Ease of use: Users can access files through interfaces that are easy to use.
You’ll find the same familiar hierarchy and intuitive folder-based organization you’re used to.

 

Let us summarize the various features in the table below:

Feature Object Storage Block Storage File Storage
Data unit Object (data + metadata) Block File
Architecture Flat (no hierarchy) Structured in blocks Hierarchical (folders and files)
Access protocol API, S3 iSCSI, Fibre Channel SMB, NFS, FTP
Scalability Unlimited Limited by hardware Limited
Performance Good for parallel loads Good for low latencies Good for sequential access
Use cases Backup, long-term storage termine Database, VM, critical applications File sharing, home directory

 

I also report in graphics the comparison between storage types and the rating (1 to 5)

With that in mind, let’s explore why object storage on-prem is crucial for backup.

USE CASES:

An e-commerce company handles millions of orders and transactions every day. They need storage that can grow along with their sales and associated data.

Or a company in the healthcare industry that needs to store patient data for 10 years, as regulations require.

Here are some data points to think about:

On average, business data grows by 20-30% each year.
Backup retention time: 15 days plus a three-year history.
We need to be able to recover data that’s been corrupted in transactions at least once a day
You can access the backup data 24/7/365.
Security in case of a ransomware attack

How well it works:

Object Storage can cluster nodes without any down time. Even if hardware fails, the platform keeps working.

If you use backup target storage for hundreds of TB or more PB, block storage isn’t as flexible when it comes to making changes, scaling out capacity, and scaling out performance. If you want to change any of these things, you’ll have to create volumes, add them to servers, format drives, and decide what you want to move. It’s a lot of work! With object storage, on the other hand, you can safely store data without having to buy new infrastructure because of hardware obsolescence.

Putting Object Storage together with backup software lets you set up specific storage policies for your most important data. Plus, you can use object lock functionality to keep your data safe before the expiration of the policy retention period—this is partly thanks to encryption, which helps reduce the risk of breaches.

Object restores are also faster than other storage types, especially compared to deduplication, and the ability to perform instant restores is really useful with object storage.

Good news! All object storage now has secure access with MFA integration. This meets one of the requirements of national or European regulations such as GDPR, NIS2 and HIPAA.

On top of that, companies also have to be able to show auditors that their data is stored properly. With object storage and backup software together, you can create detailed access and retention reports.

 

Costs and total cost of ownership (TCO)

But what’s the impact of on-premises object storage on TCO? I’ve put together a table below with some analyses that have found the reported values in a very summarized way.

Initial Costs Initial cost per TB: 30-40% higher than NAS, but offset over time. Average ROI: Within 12-24 months.
Operating Costs Reduction in energy costs: Up to 50%. Savings on hardware upgrades: 70%.
Return on Investment (ROI) Downtime reduction: 95%. Increased IT team productivity: Up to 30%.
     

So, what can we take away from all this?

On-prem object storage is a pretty innovative solution for backup data protection. It’s got some unique features that set it apart from traditional block storage. Let’s dive into some of the key points and then take a look at a chart that breaks down the differences between the two.

You can scale it up as much as you need to.
How long your data will last
Cost-effectiveness
Plus, it has advanced security features.
How well it works in the real world

 

You should think about using object storage for your data backup and protection needs today.

If you’d like more details or want to chat about the best solution for your setup, just give me a shout. I can set up a proof of concept with a range of object storage solutions and show you how they work and how they fit with some of the most widely used backup software in the world.

Interesting article Antonio and good read.  We use object storage for customers so it is on-premise in our datacenters.  😁

 
 
 

Hey ​@Chris.Childerhose what is your experience with Object Storage on prem instead other storages?

Thanks


Hey ​@Chris.Childerhose what is your experience with Object Storage on prem instead other storages?

Thanks

We use a few on premises object storage vendors the latest being Pure.  We use it on prem as we sell it to customers too.


Comment