Storage Across the Three Clouds

Three kinds of storage. Three tiers. Three providers. You've built up the concepts — block, file, object, hot, cool, archive — and now this topic lines up the names each cloud uses so any provider's documentation is immediately readable.

The concepts don't change between providers. Block storage is block storage everywhere: a virtual disk attached to one machine. Object storage works the same way on AWS, Google Cloud, and Azure. What changes is the product name and, in a few places, a fine-print detail worth knowing. Most of it is just labeling.

The Storage-Type Mapping

All three providers offer block, file, and object storage. Block storage attaches as a raw virtual disk to a virtual machine. File storage is a managed shared folder that many machines can mount. Object storage is a scalable bucket-and-object pool reached over the web. The comp-rows below list the product name each cloud uses for each type.

Tier Names Compared

The hot/cool/archive ladder exists on all three clouds, but each uses its own labels. AWS groups them as S3 storage classes: Standard for hot, Standard-IA (infrequent access) for cool, and Glacier for archive. Google Cloud Storage uses Standard, Nearline, Coldline, and Archive — four steps rather than three. Azure Blob Storage uses Hot, Cool, Cold, and Archive.

The logic is the same on all three: the further down the ladder, the lower the per-gigabyte storage cost and the higher (or slower) the retrieval. Only the names differ.

Small Differences That Matter

A few practical details differ across providers, and they show up in billing.

Minimum storage durations. Cold and archive tiers on all three providers enforce a minimum billable period — for example, 30 days for a cool tier, 90 or 180 days for archive. If you move or delete data before that period, you're charged for the remainder. The exact numbers vary by tier and provider; check the current pricing page before committing large amounts of data to a cold tier.

Retrieval fees. Some tiers charge a per-gigabyte fee to read data back, on top of the storage charge. This is intentional: cheap storage partly offsets its cost by charging you when you retrieve data rarely, which is the expected usage pattern.

Google's extra step. Google splits the "cool" region into Nearline (read roughly once a month) and Coldline (read roughly once a quarter), giving one more pricing rung between Standard and Archive compared to AWS's two-tier approach.

What Stays Constant

Across all providers: buckets hold objects, block disks attach to virtual machines, file shares mount across machines, and the tier ladder trades access speed for lower cost. Once you understand those models, you can read any provider's storage documentation and recognize every piece.