I85PREP
⌘ K
Dell DCA-ISM study materials
Dell DCA-ISM study resources

Dell DCA-ISM: RAID and Storage Networking Technologies (FC, IP SAN, NAS)

Dell DCA-ISM · Work through your exam, one topic at a time.

All guide sections

RAID and Storage Networking Technologies (FC, IP SAN, NAS)

This domain carries the most arithmetic on the exam: RAID geometry, usable capacity and write penalty, plus the transport and protocol details that separate FC SAN, IP SAN and NAS.

RAID essentials

  • Terms: strip, stripe, element (chunk) size, parity, hot spare, rebuild, expansion. Parity is distributed across all drives in RAID 5 and 6.
  • RAID 0 stripes with no protection (capacity n, penalty 1). RAID 1 mirrors (capacity n/2, penalty 2). RAID 5 has single parity, capacity n-1, tolerates one drive failure, penalty 4. RAID 6 has dual parity, capacity n-2, tolerates two failures, penalty 6. RAID 10 mirrors then stripes, capacity n/2, penalty 2, fast rebuild.
  • Large NL-SAS drives favour RAID 6 because rebuild is slow and a second failure during rebuild is realistic. Write-heavy databases favour RAID 10 despite the capacity cost.
  • A dedicated hot spare is bound to one array; a global hot spare can serve any array in the domain. Rebuild time scales with drive size and host load, not with RAID level alone.

Fibre Channel SAN

  • Switched fabric versus arbitrated loop. Login sequence is FLOGI to the fabric, then PLOGI to the target port; identity uses WWNN and WWPN.
  • Zoning by WWN survives a host HBA replacement in the same server role; zoning by port or domain does not. Single-initiator, single-target zoning limits blast radius. LUN masking on the array is a separate control from fabric zoning.
  • Current link speeds are 16, 32 and 64GFC; NPIV presents multiple virtual WWPNs through one physical port; FC-NVMe shares the fabric with a different upper-layer protocol.

IP SAN and NAS

  • iSCSI carries SCSI over TCP port 3260, names targets and initiators with IQNs, discovers with SendTargets or iSNS, authenticates with CHAP, and relies on MPIO for path redundancy. Jumbo frames must be consistent end to end or performance falls rather than fails.
  • NAS is file-level: NFS v3 is stateless with separate locking, NFS v4 is stateful with integrated locking and ACLs. SMB 3.x adds multichannel, SMB Direct over RDMA and continuous availability for transparent failover.

Worked example

Twelve 10 TB drives in RAID 6 with one global hot spare: the spare leaves 11 drives in the RAID set, usable capacity is (11 - 2) times 10 TB, or 90 TB, from 120 TB raw, about 75% efficiency. If the volume is thin provisioned for 90 TB but only 40 TB is written, physical consumption stays near 40 TB plus parity overhead.

Traps that cost marks

  • Omitting the hot spare from capacity math.
  • Choosing parity RAID for sustained small random writes the penalty will crush.
  • Confusing zoning with LUN masking; both are required for host isolation.