Skip to main content

Keys to Understanding and Choosing Your Backup as a Service (BaaS)

· 3 min read
Malo Paletou
Founder of Datashelter

A quick Google search reveals that the term BaaS is used across all sectors. For instance, you might find Banking as a Service or Backend as a Service.

However, the topic we aim to understand today is Backup as a Service. Some might call it "the BaaS for bearded folks who manage their own servers"

What Does Backup as a Service (BaaS) Mean?

Backup as a Service involves offering a backup service through a cloud model. Instead of having to provision infrastructure and storage dedicated to this purpose, backup is consumed on demand.

Typically, this service operates on a "pay-as-you-go" basis. For example, costs might be based on the number of servers backed up or the amount of storage consumed.

Three Categories of Backup Solutions

Various backup solutions are available in the market. We can distinguish them into three categories:

  • On-premise appliance installation: This could be a specific machine designed to write backups on tape, or a simple NAS when that suffices.
  • Backup software installation on the client’s infrastructure, as offered by companies like Veeam.
  • Cloud-based "as a Service" solutions (BaaS) that eliminate the need to install a dedicated backup component in your infrastructure. This is particularly suitable for the needs of small and medium-sized businesses.

Installing a complex backup appliance is justified for large companies managing a vast private cloud with thousands of servers. However, this approach can be overkill for small businesses or those that rent just a few servers from a cloud provider for more limited needs, such as hosting business applications or even just a website.

For this more common use case, Backup as a Service solutions are particularly well-suited. They are generally set up in just a few minutes, and allow you to benefit from professional backups at a relatively low monthly cost.

How to Choose Your Backup as a Service?

The market is full of Backup as a Service (BaaS) offerings. However, there are several important considerations to keep in mind to make the right choice:

  • Backup Methods: How does the provider execute backups? Are they performed incrementally, with compression, or through deduplication?
  • Access Requirements: What specific access rights does your backup provider require? Do I need to grant specific accesses to my servers ?
  • Notifications and Alerts: Will you be notified in case of errors during the backup process? How ?
  • Storage and Security: Where and how will your backups be stored? Determine whether encryption is handled on the client side.

Thinking about each of these considerations will help you feel more confident as you make your decision.

About Us

At Datashelter, we are committed to offering you the best in data backup solutions with just a few clicks, through reasonable and predictable pricing.

If you are looking for the ultimate backup solution, we invite you to visit datashelter.tech and challenge our team of experts who are ready to support you at every step of your project!

The Future of Secure Backups: A Deep Dive into Datashelter’s Unique Security Features

· 3 min read
Malo Paletou
Founder of Datashelter

At Datashelter, security is at the heart of our backup solutions. That's why we have designed our product to offer strong, auditable guarantees that set us apart in the market for outsourced backup services.

This article will detail the main features Datashelter offers to ensure the security and integrity of your backups.

No SSH Access Required

Firstly, we've taken a different approach compared to most of our competitors. While many competing solutions require SSH access to the machine to schedule your backups, this will never be requested at Datashelter.

Thanks to the control we have over our CLI tool, snaper, we've implemented a push-only model. This means that your servers push (or pull) data to/from Datashelter, but the reverse is not possible. In addition to ensuring maximum security "by design" this also means you don't need to configure special permissions on your firewall to perform your backups.

Furthermore, all communications between your servers and Datashelter are conducted through an S3-compatible API encrypted with TLS. In summary, you do not need to give us access to your server to perform your backups!

AES-256 Encryption

With our in-house tool, snaper, we have implemented native data protection features. This includes AES-256 encryption of your data. Your data is encrypted with AES-256 using your own key before even being sent to us.

So, we are not misleading you when we say that not even Datashelter can access your data.

Maximum Access Segmentation

Moreover, we have generalized an essential principle of security right from the stage of creating your servers. Indeed, we generate a pair of S3 credentials (access & secret key) and a dedicated bucket for each of your servers configured on our platform.

This significantly reduces the potential attack surface in case one of your machines is compromised.

Immutable S3 Storage

Lastly, I must tell you about how we store your data at Datashelter.

We have developed a unique technology that prevents any alteration of your previous backups. We call this immutable S3 storage.

You may have noticed, your servers communicate with Datashelter through a custom S3 gateway. This gateway allows us to implement additional security rules.

We can mention two of them here:

  • Deletion request blocking: you do not have the ability to delete your backups directly. Only Datashelter lifecycle rules have the ability to do so.
  • Prevention against object rewriting: our gateway exposes a standard S3 API, while we use versioned S3 buckets in the backend. So if you try to rewrite data (you or an attacker who has seized your S3 credentials), we will accept the request but we also retain the previous version of that object (which, as a reminder, cannot be deleted).

In this way, you have the ability to restore your backups even in the event of a compromise of the credentials used for your backups.

With this step now complete, your data is finally stored on our infrastructure. We currently rely on the infrastructure of our partner OVHcloud to offer you multi-certified storage (ISO 27011, HDS & soon SecNumCloud).

Your data remains safe at Datashelter, and will be there when you need it most!

Understanding Data Backup: The Importance of Structured and Unstructured Data

· 2 min read
Malo Paletou
Founder of Datashelter

If you're new to data backup, the terms "structured" and "unstructured data" might be unfamiliar. These concepts are crucial in selecting the right backup solution. Let's explore these terms in a straightforward, jargon-free manner.

What is Structured Data?

Structured data refers to information formatted in a way that is easily interpretable by the software used for backing it up. This means the backup format is directly linked to the application being backed up.

For example, consider two methods of backing up a relational database like MySQL or PostgreSQL:

  • Performing a SQL dump, which is an example of backing up structured data.
  • Copying the entire data storage directory, which would be considered backing up unstructured data.

Choosing to backup structured data means your data is exported in a way that maintains the consistency and integrity of the information, ensuring the backup is restorable and reliable.

Challenges with Unstructured Data Backup

On the other hand, backing up unstructured data, although simpler (often involving a mere copy of the data directory), carries risks such as potential corruption or data being unreadable and unrestorable.

This risk becomes apparent when taking snapshots from cloud services or using automatic backup features on solutions like an OVH VPS.

Conclusion: A Balanced Backup Strategy

While structured data backup is our preferred method at Datashelter—guaranteeing restorability—it's clear that a nuanced approach, combining both structured and unstructured backups, can be beneficial.

Understanding the distinction and application of structured versus unstructured data backups will ensure you're better prepared to protect your data effectively, reinforcing the adage that a backup isn’t truly a backup unless it's restorable.

Exploring Backup Strategies: From Traditional to Index-Based Solutions

· 2 min read
Malo Paletou
Founder of Datashelter

Backups are a crucial topic in any tech-driven workplace, whether you're a system administrator or the developer tasked with maintaining servers. Understanding the various backup typologies and their implications is essential, and this article aims to clarify these types and introduce a new paradigm in backup solutions.

Understanding Traditional Backup Typologies

Traditional backup methods have been extensively discussed, yet they remain foundational. There are three primary typologies you'll encounter:

  • Full Backup: This method duplicates the entire production medium onto a second storage medium. It's straightforward and quick to restore but requires significant storage space.

  • Incremental Backup: This starts with a full backup and only adds files that have been modified since the last backup. While it saves on storage space, the restoration process can be complex as it requires all previous incremental backups.

  • Differential Backup: Similar to incremental, this method also begins with a full backup but adds all files changed since the last full backup. Differential backups are easier to restore than incremental ones, requiring only the last full backup and the latest differential backup.

Introducing Index-Based Backup

While traditional methods are well-known, index-based backup is gaining traction. This innovative approach relies on maintaining an index to restore data, which can be a slight drawback if the index format is proprietary or complex. However, its significant advantage lies in the ability to restore deleted files efficiently.

Technologies such as restic, borgbackup, snaper (Datashelter), and Synology HyperBackup are pioneering this method. With the rise of object storage, index-based backups are becoming more feasible and scalable, with no limits on file numbers or bucket sizes. Data in an index-based system might be organized as follows:

  • /data: Contains file blobs representing complete files or data blocks.
  • /index: Stores indices like index_1, index_2, etc.

Conclusion: Comparing Backup Typologies

Each backup typology offers unique advantages and challenges. The table below summarizes these to help you choose the most suitable backup strategy for your needs.

Storage requirementsBackup timeEase of restorationConsideration of deletionsInterdependence between backups
Full++++++++YesNone
Incremental+++NoTotal
Differential++++NoPartial
Index based+++++YesNone

Why Datashelter?

· 3 min read
Malo Paletou
Founder of Datashelter

On March 10, 2021, one of OVH's Strasbourg data centers caught fire, taking 3.6 million websites with it. Many web developers, consultants and agencies remembered, and the 8 o'clock news made it the lead story.

At the time, I was an infrastructure consultant. Initially awakened by on-call calls, I soon found myself inundated with calls from my customers: Malo, how long is this going to last? Can we restore our sites elsewhere? Malo, what should we tell our customers?

That day, two subjects took on particular importance: the location of backups and the ease with which they can be restored.

For all too many companies who put their blind trust in their hosting provider to create and maintain their backups, this incident proved critical, if not fatal.

Some companies were able to restore a backup with a delay of several days, while others lost all their data. These were not backups, but snapshots, stored on the same server as their production.

This example perfectly illustrates two of my main concerns today:

  • off-site backups are essential
  • a backup that can't be restored isn't a backup

Data Backup is Not A Sexy Subject

Fortunately, my customers had backups to restore 😜. But this incident got me thinking about why this type of incident has hurt so many companies. Why didn't so many companies have resilient backup processes in place?

In my opinion, there are two reasons explaining this situation:

  • major obstacles to implementation
  • complexity of the procedure

To begin with, complexity. What software should I use? What backup policy should I implement? How will I restore my data?

Complexity can never be eliminated, but it can be abstracted to provide a simple, intuitive experience for the user. Like Qonto in banking or Alan in insurance.

Secondly, when you start thinking about corporate backups, you quickly come up against a number of obstacles. Where should I store my data? Do I need to back up structured or unstructured data? How much will it cost?

We usually end up adopting one of these 3 solutions:

  • develop a little python/bash script to do the bare minimum
  • pay for expensive licenses from Veeam or Acronis, where the price is never clear
  • ignore the problem or put it off until later

Datashelter is Here

That's why I created Datashelter. We want to offer a turnkey, sovereign data backup automation solution for small and medium-sized businesses.

This complete solution is built around a simple, predictable price (€2/connected server + €5/TB) and, in just a few clicks, enables you to have resilient backups, be alerted in the event of an anomaly, and be supported when you need it most.

Quite simply, we want to be a no-brainer when it comes to securing your company's data at the lowest possible cost. Whether you're renting a small VPS or hosting hundreds of applications with the most demanding availability constraints.

We believe Datashelter is the tool that makes your life easier. So we're giving you 30 days to try it out, and set up your first backups for free