Storage
What to know
OCI storage splits into block, file, object, and archive. Block volumes are network-attached disks for instances: boot volumes and additional data volumes, with performance tiers driven by VPUs (lower cost, balanced, higher performance, ultra-high performance) plus backups, clones, cross-region replication, and cross-AD replication. File Storage provides NFS shares through mount targets inside a subnet. Object Storage provides buckets inside a namespace, with tiers (Standard, Infrequent Access, Archive), versioning, lifecycle rules, pre-authenticated requests, and replication.
The key skill is matching durability, access pattern, and cost. Block volumes attach to one instance at a time unless you deliberately use multi-attach with a cluster-aware file system. Archive is cheapest but has retrieval latency and a minimum storage duration; Infrequent Access also has a minimum retention before early-deletion charges.
Traps
- Assuming a block volume can be shared by several instances by default.
- Forgetting that lifecycle rules act on objects by age, prefix, or pattern, and that transitions move only toward colder tiers.
- Using Archive for data that must be read frequently or immediately.
- Confusing the tenant-wide bucket namespace with compartment placement.
Worked example
Logs written continuously and read only during incidents: store in Object Storage Standard, add a lifecycle rule to move objects to Infrequent Access after 30 days and to Archive after 90, then delete after 365. Place the bucket in a compartment whose policy grants auditors read objects only.