What it is
S3 storage classes are different price and access profiles for the same object store. Azure Blob has the same idea as access tiers: hot, cool, cold, archive. Different labels. Same shopping problem.
Frequent access costs more per gigabyte and serves quickly. Infrequent and archive classes cost less to store and more, or slower, to retrieve. Lifecycle rules can move objects automatically as they age.
The decision is not "which is best." It is "how often do we need this data, and how fast?" If nobody knows, everything stays in Standard forever, which is a choice. An expensive one.
Why it matters in the meeting
Leaving everything in Standard is simple and expensive. Parking active data in deep archive is cheap until someone needs it today and discovers a restore window. Bart's version: put data on the right shelf.
JJ hears "we moved logs to Glacier" and thinks savings. Ask how often legal needs those logs this quarter. If the answer is weekly, you bought a freezer for milk.
Real world
Log archives, compliance holds, and media libraries are classic candidates for cooler classes. Customer-facing assets and active working sets are not.
A lifecycle policy tied to real access patterns usually pays for itself faster than a one-time cleanup project. AWS and Azure both offer the policy. Someone still has to write it and mean it.
Storage class is a business decision about access frequency dressed up as a dropdown. Hot for work. Cool for maybe. Archive for "wake me if the auditors call."
What to ask
- What share of our object storage is in the default hot tier with no lifecycle rule?
- Which data do we need in minutes versus days, and is that written down?
- Have we tested a restore from archive, or only celebrated the cheaper storage line?
- Who owns the lifecycle policy, and when was it last reviewed against real access?
You just knew a little more Jack than you did five minutes ago.
All concepts