NIYYASOFT

Azure Local — Storage Sizer

Capacity planning with Microsoft Learn published rules: resiliency efficiency, reserve capacity, cache, nested resiliency and platform limits
1. Platform
2. Capacity drives (per server)

Enter sizes in decimal TB as sold by the vendor. Only capacity drives count toward raw capacity.

3. Cache drives (per server)

The fastest drive type becomes cache automatically. Required for HDD capacity; not supported on single machines.

4. Volume resiliency

5. Reserve & system volumes

Azure Local creates a 250 GB infrastructure volume. Performance history uses a ~10 GB volume. Both are mirrored by server count.

6. Workload requirement (optional)

Data reduction (deduplication/compression) is your own estimate. Leave it at 0 for a conservative, Microsoft-rule-only result.

Where the raw capacity goes

Cache & memory

Proposed volumes

PowerShell (run on a cluster node after review)


      

    

    

      
Usable capacity for every resiliency type on the same hardware, after reserve and system volumes. The highlighted row is your current selection.

    

    

    

      

        
How the sizing is calculated

        
Every rule comes from Microsoft Learn documentation for Storage Spaces Direct, the storage layer of Azure Local and Windows Server clusters. Niyyasoft isn't affiliated with Microsoft; confirm final hardware with your Azure Local solution provider and the Azure Local catalog.

        
    
          
  1. Raw capacity = servers × capacity drives × drive size. Cache drives don't add usable capacity.
    2. 
          
  2. Reserve: leave one capacity drive per server unallocated, up to 4 drives, for in-place repair.
    3. 
          
  3. System volumes: infrastructure volume (250 GB) and performance history (~10 GB), two-way mirror on 1–2 servers, three-way on 3+.
    4. 
          
  4. Usable = (raw − reserve − system footprint) × efficiency: two-way mirror 50%, three-way 33.3%, dual parity 50–80% by fault domains (hybrid 4–6 RS 2+2 · 7–11 RS 4+2 · 12–16 LRC(8,2,1); all-flash 4–6 RS 2+2 · 7–8 RS 4+2 · 9–15 RS 6+2 · 16 LRC(12,2,1)), nested two-way 25%, nested MAP by Microsoft's formula.
    5. 
          
  5. Volumes: at least one per server, in multiples of the server count, ≤ 64 TB each, ≤ 64 per cluster.
    6. 
          
  6. Cache: ≥ 2 cache drives per server (≥ 32 GB); capacity drives a multiple of cache drives; Azure Local hybrid ≥ 15% cache-to-capacity; 4 GB RAM per TB of cache.
    7. 
          
  7. Limits: 400 TB raw per server, 4 PB per pool; Azure Local data drives ≥ 500 GB. Units: decimal TB in, binary TiB as Windows shows it.
    8. 
        

        
Fault tolerance · Plan volumes · Nested resiliency · Cache · Azure Local requirements