Storage tiers
Also known as: storage tiering, hot and cold storage, tiered storage, cloud storage tiers
Storage tiers are levels of storage with different speeds and prices, matched to how often data is used. 'Hot' storage holds frequently accessed data on fast drives and costs more. 'Cool' or 'infrequent access' tiers cost less but charge more to retrieve data. 'Archive' tiers, such as Amazon S3 Glacier Deep Archive or Azure Archive, are extremely cheap for storage but can take hours to retrieve from.
While tiering can cut storage costs dramatically, retrieval fees and delays can surprise people during urgent restores.
Storage tiers, explained at three levels
At home
Storage tiers are like keeping everyday clothes in your closet, seasonal clothes in the basement, and old keepsakes in a storage unit across town.
At work
Moving rarely used data to cheaper tiers can significantly reduce cloud storage bills.
At home
Some photo services keep recent photos instantly available and compress or archive older ones. Check how long it takes to get originals back.
At work
Cloud tiers differ in storage price, retrieval cost, retrieval time, and minimum storage duration. Lifecycle rules can move data automatically after a set number of days.
At home
Cheap archive storage can be expensive to retrieve all at once. In my opinion, know the restore cost before trusting it with your only copy.
At work
Classify data by access patterns, set lifecycle policies, watch minimum duration charges and early deletion fees, model restore costs and times against RTOs, and test archive retrievals.
Real-world examples
At home
Thomas archived years of raw video footage to a cheap cloud archive tier, cutting his bill, but noted it would take hours to download if needed.
At work
Lena manages cloud storage for a Toronto architecture firm. Lifecycle rules move project files older than two years to an archive tier, saving thousands annually.
How it compares
| feature | Hot | Cool | Archive |
|---|---|---|---|
| Storage cost | Highest | Lower | Lowest |
| Retrieval time | Instant | Instant | Hours |
| Retrieval fees | None or low | Moderate | Higher |
| Typical data | Active files | Older backups | Long-term records |
Colder tiers are cheaper to keep and costlier, slower to retrieve.
For analysts
In storage cost reviews, I'd map data by how often it's actually used. Many organizations pay hot-storage prices for files nobody has opened in years.
History & origin
Hierarchical storage management, moving older data from expensive disks to cheaper tape, was used on mainframes from the 1970s. As cloud storage grew, Amazon launched Glacier in 2012 for low-cost archiving, followed by infrequent access tiers and S3 Glacier Deep Archive in 2019, priced around $1 US per terabyte per month. Microsoft and Google added comparable tiers. Automatic tiering features, such as Amazon's S3 Intelligent-Tiering in 2018, moved data between tiers based on use. In my view, storage tiers are one of the easiest cloud cost savings, as long as you plan for how and when data comes back.