RAID and Storage Operations
What this domain tests
RAID and Storage Operations tests controller choice, virtual disk design, rebuild and foreign configuration handling, and storage configuration through iDRAC, Lifecycle Controller, or OpenManage. Expect capacity and fault-tolerance reasoning, not just RAID level definitions.
Core rules
- PERC controllers provide hardware RAID with cache and battery or capacitor backup. HBA mode passes drives directly to software-defined storage such as vSAN.
- BOSS provides a dedicated boot device, usually two M.2 drives in RAID 1, keeping the OS off data disks.
- RAID 0 has no redundancy; RAID 1 mirrors; RAID 5 tolerates one drive failure; RAID 6 tolerates two; RAID 10 mirrors then stripes.
- A hot spare is a standby drive that rebuilds automatically. Confirm it is assigned to the right virtual disk and is large enough.
- Patrol read and consistency checks find latent media errors. They are not the same as a rebuild.
- Foreign configuration occurs when a controller sees disks from another controller. Import it only when the old configuration is known and wanted.
- Self-encrypting drives require key management through the controller; losing keys means losing data.
Worked example
Four 4TB drives are available for a RAID 5 data volume, and one additional 4TB drive must be a hot spare. The design uses all five drives: four in RAID 5 plus one spare. Usable capacity is approximately three drives, because RAID 5 loses one drive to parity. The reasoning is that the spare is not counted as capacity and RAID 5 tolerates only one failure; adding the spare allows automatic rebuild.
Traps
- Assuming RAID 5 is safe for large drives without considering rebuild time and URE risk.
- Mixing SAS and SATA drives in one virtual disk, or mixing sizes in a way that wastes capacity.
- Importing a foreign configuration blindly.
- Forgetting that a rebuild degrades performance.
- Using hardware RAID for a platform that requires HBA or direct-attached disks.