Storage and Data Transfer Options
What this domain tests
Matching data to the right OCI storage service and the right transfer path - online versus offline - including the bandwidth arithmetic that decides between them.
Core rules and concepts
- Object Storage holds unstructured data, backups, and migration staging; tiers are Standard, Infrequent Access, and Archive, with lifecycle policies moving data between them.
- Block Volume backs VM boot and data disks with low latency per instance; File Storage provides shared NFS mount targets for clustered or shared-access workloads.
- Online transfer: OCI CLI, rclone, or Transfer Manager with multipart parallel uploads over the internet, or FastConnect/VPN when traffic must stay private or needs sustained throughput.
- Offline transfer: OCI Data Transfer Service ships encrypted appliances to your site; you load data, ship back, and OCI ingests into Object Storage or Archive Storage. Choose it when link math says months, not weeks.
- Always verify integrity: checksums, object counts, and sizes after any bulk transfer.
Worked example
500 TB must reach Object Storage. A dedicated 1 Gbps circuit delivers about 125 MB/s theoretical, roughly 450 GB/hour, so 500 TB is about 1,100 hours - around 46 days at perfect efficiency, nearer 65 days at realistic 70%. That exceeds the project window, so the answer is Data Transfer Service: stage onto appliances, ship, and ingest, while a FastConnect circuit carries ongoing incremental changes. The deciding factor is arithmetic, not preference.
Traps that cost marks
- Assuming a 1 Gbps circuit is symmetric and fully usable for a single TCP stream.
- Using Archive Storage for data that must be read back immediately after migration.
- Skipping multipart uploads and parallelism, then blaming the network.
- No checksum reconciliation, so silent corruption surfaces in production.
- Forgetting that Block Volume is per-AZ replicated, which affects cross-AZ migration design.