Cloud Storage and Data Management
What to know
Match storage type to access pattern. Object storage is for unstructured data at scale, accessed via API, with durability and lifecycle policies; block storage attaches to a single compute instance like a disk; file storage provides shared hierarchical access across many clients. Know durability (probability data is not lost) versus availability (probability it is reachable) - object stores often advertise very high durability with lower availability SLAs. Understand redundancy options (single-AZ vs. multi-AZ vs. cross-region replication) and their cost/latency trade-offs. Lifecycle policies move data to colder tiers or delete it. Consistency models matter: strong consistency means reads reflect the latest write.
Data management
Know backup versus archive, retention and legal hold, and data classification. Caching improves read performance but introduces staleness.
Traps
Questions swap durability and availability, or imply object storage is a good fit for a database's live files (it is not - use block storage). Don't assume cross-region replication is free or instant. Lifecycle transitions have minimum-duration charges in cold tiers, so aggressive tiering can cost more, not less.
Worked example
A company stores years of compliance records rarely accessed but must retain for seven years. Answer: object storage with lifecycle transition to an archive tier and a retention policy - not block storage, and not deletion. Rule of thumb: choose storage by access pattern and sharing model, then tune durability and tiering to the requirement.