Where database internals are a fair question rather than a bonus - storage, transactions, and OCI-scale infrastructure.
Oracle is one of the few places where database internals are squarely on the table. Depending on the team, you may be asked how a B-tree index behaves under heavy insert load, what a write-ahead log buys you during recovery, how MVCC avoids read locks, or why a query planner picked a particular join. Most candidates prepare databases as a usage topic and are caught out by the depth.
The OCI side asks infrastructure design: multi-tenant compute, storage durability guarantees, control plane versus data plane separation, and regional isolation. Both sides share a bias toward correctness and durability over novelty, which is consistent with the products - customers run their systems of record on them.
45 min
Algorithms with complexity reasoning.
45-60 min
Two rounds, implementation quality weighted.
60 min
Storage engines, transactions, indexing, or cloud infrastructure depending on the team.
60 min
Service design with durability and isolation as first-class constraints.
45 min
Ownership and working within long-lived, correctness-critical codebases.
Genuinely deep for database and storage teams. Index structure behaviour under load, recovery via write-ahead logging, MVCC, and query planning are fair questions rather than bonus material.
Yes. OCI roles ask cloud infrastructure design - multi-tenant compute, storage durability, control plane separation, regional isolation - while database roles go into engine internals. Confirm the team before you prepare.
Medium difficulty and weighted on correctness and clarity rather than speed. The deciding signal is usually the systems or design round.
Practise the rounds Oracle actually runs, and get scored on where you stand.