No one makes the decision to store more. It just happens. Logs are saved just in case, old projects are left as they are, and every year a bit more disk space is purchased. Until someone does the math and discovers that storage has become one of the larger line items in the IT budget.
The problem is rarely the quantity.
It is that everything is treated the same. Data used daily and data that no one has opened in four years sit on the same fast, expensive storage, because that was easiest when it was created.
In most environments, only a small portion of the data is actively used. The rest simply needs to remain—for legal reasons, as a precaution, or because no one has dared to decide otherwise.
Distinguish between what is active and what is dormant
Active data demands performance: databases, files in daily use, virtual machines. Fast storage on a SAN or equivalent is the right investment here.
Dormant data only needs to be accessible when requested. Here, cold storage is significantly cheaper per stored unit, with the tradeoff that retrieval takes longer and incurs some cost.
Unstructured data — images, documents, logs, sensor readings — usually belongs in object storage, which scales without requiring anyone to plan volumes in advance. It’s also the foundation once volumes start approaching big data levels.
The cost you don’t see until you need to move
Many cloud providers charge you to get your own data out. As long as your data just keeps growing, you won’t notice. You’ll notice the day you switch providers or need to restore a large amount — and that’s when the egress cost can determine whether the switch actually happens.
It’s worth checking the contract before the data volume grows large enough to make the question expensive. FinOps is fundamentally about just that: knowing what you’re paying for before the bill arrives.
Start by finding out what you actually have
A review of access patterns usually gives a clear picture: what’s used daily, what’s used occasionally, and what hasn’t been touched since it was created.
Only then is it possible to put a structure in place that holds up — which data should go where, how long it should be retained, and when it moves between tiers. Without that structure, every new storage investment just extends the same problem.
Here’s how we do it at Savecore
Savecore S3 is object storage in our own data centers in Sweden, with a pricing model you can calculate in advance. We publish our package prices openly.
If you need to separate active and archived data, we can help you set up a structure for it, so you only pay performance-tier pricing for the data that actually needs performance. If the storage is part of a larger infrastructure, it’s combined with our IaaS platform.
Frequently asked questions about growing data volumes
How do we know what’s active data and what’s dormant data?
Access patterns reveal it. A review usually shows that a large share of the data hasn’t been touched in several years.
What does it cost to retrieve data again?
Our pricing model is transparent and published. Check the same with your current provider — that is usually where the unpleasant surprises lurk.
Can we migrate legacy data without disrupting production?
Yes. Archiving inactive data is typically done without impacting running systems.
Pay performance pricing only for data that actually requires performance.
We review what you store and what it should cost.