Tier 0 Metadata Tier Local Drive Redundancy (D+x)

[jasnoyaeger]

Summary

Document how the Tier 0 metadata tier mirrors metadata across a node's local drives (D+x) and how that copy count follows the system redundancy level.

Type

Conceptual

Suggested Content

• Audience: Admins sizing or deploying controller nodes
• Prerequisites: Understanding vSAN Redundancy Levels
• Key sections:
  • Metadata matches system resilience locally: N+1 → 2 local copies (D+1), N+2 → 3 local copies (D+2)
  • Copy count is capped by T0 drives present per node: 1 drive → 1 copy, 2 drives → 2 copies, 3+ drives → 3 copies
  • Usable T0 capacity ≈ one drive's size; additional drives hold mirror copies (e.g. 2 × 1.6TB → ~1.6TB usable)
  • Sizing implication: N+2 systems need 3 T0 drives per controller node for metadata resilience to match the system level
  • Suggested location: extend the vSAN Redundancy Levels page or the Tier 0 section of Storage Tiers

Context

Requested via support interaction. Behavior confirmed by engineering; current docs cover only cross-node copy counts (N+1/N+2), not the metadata tier's local drive mirroring.