🔗 Oracle Database Administration – Oracle RAC Architecture & Cache Fusion
20 questions · 15 min timer · Explanations included · Free
RAC is where Oracle stops being a database on a server and becomes a distributed system — with all the coordination costs that implies. Cache Fusion is the mechanism at its centre: shipping blocks between instances over the interconnect instead of forcing a disk round trip.
Most RAC performance problems trace back to that interconnect, and most RAC interview questions are checking whether you know why.
What this quiz covers
- Cache Fusion — block transfer between instances and how consistency is maintained
- GES and GCS — the Global Enqueue Service, the Global Cache Service, and the global resource directory
- Clusterware — node membership, heartbeats, and split-brain resolution
- SCAN and VIPs — client connection routing and what happens on node failure
- The interconnect — the usual causes of contention and how it should be configured
- Node failure — instance recovery and redo synchronisation across instances
- Rolling patches and RAC One Node
- Virtualised RAC — the configuration decisions that matter
Common mistakes
- Treating the interconnect as ordinary network plumbing rather than the component that determines RAC performance
- Confusing SCAN with a VIP; they solve different problems
- Assuming Cache Fusion removes disk I/O rather than shifting where the cost lands
Twenty questions below, each with an explanation.
