Object storage is the most common way to keep growing data such as backups, archives, website images and logs cheaply and at scale. We explain how it differs from block and file storage, the S3 API, versioning, object lock and the cost model in plain language.
Company data grows every year: accounting backups, camera recordings, product images, scanned contracts, application logs. Most of it rarely changes, but it has to be kept for a long time. Squeezing it onto server disks or a file server quickly becomes expensive and hard to manage. Object storage was designed for exactly this kind of data.
In this article we explain what object storage is, how it differs from block and file storage, what "S3-compatible" actually means and which jobs it is the right choice for, without drowning you in jargon.
What Is Object Storage?
In object storage, data is not kept in a folder tree or in disk blocks but as independent units called objects. Each object has three parts:
- The data itself: a PDF, a photo, a backup file, a video or a log archive.
- A key: the object's unique name, for example invoices/2026/october/invoice-1045.pdf. The slashes look like folders, but they are simply part of the name.
- Metadata: extra information describing the object, such as file type, creation date or custom tags you add, like "department: finance".
Objects live in containers called buckets. Think of a bucket as a large store room with its own access rules and settings. You might create three buckets for "backups", "web-images" and "logs", each with different permissions and retention rules.
You do not access object storage by mapping a network drive. Instead, applications talk to it through an API over HTTPS. Your application or backup software sends requests such as "upload this object", "fetch that object" or "list the objects in this bucket".
Object vs Block vs File Storage
All three store data, but each is optimised for different work. A quick comparison:
| Feature | Block Storage | File Storage | Object Storage |
|---|---|---|---|
| What it looks like | A raw disk attached to a server | A share of folders and files | Objects in a bucket, accessed by key |
| Access method | The OS sees it as a disk | Network drive over SMB or NFS | S3 API over HTTPS |
| Typical use | Databases, virtual machine disks | Shared office folders | Backups, archives, media, logs |
| Performance | Very low latency | Moderate | High aggregate throughput, higher latency |
| Scaling | Limited by disk size | Limited by server capacity | Practically unlimited |
| Cost | Highest | Moderate | Lowest per GB |
In short: the disk your database runs on is block storage and should stay that way. Shared folders your team opens and edits every day work well on file storage. But for growing data that is kept for years and mostly written once and read occasionally, object storage is cheaper and far more flexible.
What Do "S3 API" and "S3-Compatible" Mean?
S3 is the name of Amazon's object storage service. Over the years its API became the de facto industry standard. Today most backup software, backup tools, web applications and data tools know how to talk to the S3 API.
S3-compatible storage is storage that supports this API but is not operated by Amazon. In practice, a tool written for Amazon S3 will usually work with another S3-compatible service by simply changing the endpoint address, access key and secret key. That keeps you from being locked into a single provider.
To connect to S3-compatible storage you typically need:
- An endpoint address (the HTTPS address the provider gives you)
- An access key and secret key (think of them as a username and password)
- A bucket name
- A region, if required
Compatibility is not always one hundred percent. Some advanced features, such as object lock or lifecycle rules, may not behave identically everywhere. Always test the features your software relies on with the provider you are considering.
What Is Object Storage Used For?
Backup
This is the most common use. Veeam Backup and Replication can use S3-compatible object storage as a backup repository, so you can send a copy of your local backups straight to the cloud. For Linux servers and smaller environments, open-source tools such as rclone and restic also support S3 targets. If office disks fail, a server is stolen or ransomware strikes, a copy of your backups stays safe somewhere else.
Website and Application Media
E-commerce product images, user-uploaded documents and videos fill application server disks fast. Moving these files to object storage lightens the server, simplifies backups and lets multiple servers access the same files.
Archives
Invoices, contracts, email archives and old project files with legal retention periods are a perfect fit. They are rarely accessed, but can be retrieved within minutes when needed.
Logs and Records
Firewall, web server and application logs take up a lot of space over time. Keeping the last few weeks for active analysis and sending the rest, compressed, to object storage is a common and economical approach.
Versioning
In a bucket with versioning enabled, uploading an object with the same name does not delete the old one; it is kept as a previous version. Deleting an object actually adds a "delete marker", and earlier versions can be restored.
This is a simple but effective defence against accidental overwrites and deletions. The catch: every version takes up space and is billed. It makes sense to combine versioning with a lifecycle rule that removes old versions after a set period.
Object Lock and Immutability
Ransomware no longer targets only servers; it also tries to encrypt or delete any backups it can reach. Object lock prevents an object from being deleted or modified for a period you define. Even if an administrator password is stolen, the backup cannot be deleted during that window.
There are usually two modes:
- Governance mode: users with special permissions can remove the lock. Useful for testing and transition periods.
- Compliance mode: nobody, including the provider's administrators, can delete the object until the period expires. This is true immutability.
Backup software such as Veeam can use object lock on S3-compatible storage to create immutable backups. Note that object lock usually has to be enabled when the bucket is created; turning it on later is not possible with every provider.
The Cost Model: What Do You Pay For?
An object storage bill usually has three components. Comparing prices without understanding them is misleading.
- Storage: the average amount of data stored during the month, charged per GB. This is the most visible item.
- Egress: data downloaded out of the storage. With some providers this can exceed the storage charge itself, especially during large restores.
- Requests (API calls): every upload, listing and download is a request. For backups and applications that work with millions of small files, this adds up.
Imagine having to download all of your backups during a disaster. A provider with cheap storage but expensive egress can produce the biggest bill on your worst day. Work out the total cost with a realistic restore scenario, not just the per-GB storage price.
With CloudRUX S3 object storage, ingress and API requests are free, and monthly egress up to the amount of data you store is included. Your bill stays predictable.
How to Choose Object Storage
- Where is the data kept? If you store personal data, the country it is held in matters under KVKK, Turkey's data protection law. A data centre in Turkey simplifies cross-border transfer obligations.
- Is the S3 compatibility good enough? Test it with your backup software or application. Check object lock, versioning and multipart upload support in particular.
- What is the total cost? Add up storage, egress and request charges. Consider USD, TRY or fixed-package options to manage currency risk.
- Latency and speed: data that is close to you shortens backup and restore times.
- Security: HTTPS encryption in transit, per-bucket permissions and separately managed access keys are baseline expectations.
- Support: reaching a team that speaks your language and works in your time zone is invaluable, especially during a restore.
Frequently Asked Questions
Can I use object storage like a drive on my computer?
Some tools let you mount a bucket as a drive, but object storage is not designed for that. File shares are better for office documents that change often. Object storage works best for backups, archives and application data.
Is S3-compatible storage as secure as Amazon S3?
Security depends less on the API and more on the provider's infrastructure and your configuration. With HTTPS encryption, strong access keys, per-bucket permissions and object lock, S3-compatible storage provides strong protection.
Can I send my Veeam backups directly to object storage?
Yes. Veeam supports S3-compatible object storage as a backup repository. A common setup keeps a local repository and sends a copy to object storage, which you can make immutable with object lock.
How much data should I start with?
The beauty of object storage is that you can start small and grow as needed. Begin with a few hundred GB of backup copies and move archive and media data over time.
CloudRUX S3 Object Storage
CloudRUX S3 is an S3-compatible object storage service that keeps your data in data centres in Turkey. It works with Veeam, rclone, restic and any other tool that speaks the S3 API, with free ingress and free API requests. See packages and technical details on our S3 object storage page, or get in touch to plan your backup or archive scenario with us.