Kepler-47A Front
Kepler-47A in Production @ Client Site

Our Kepler-47 SMB and Specialised 2-Node Hyper-Converged Cluster Provides Cloud-Like Resilience for those On-Premises Workloads!

The two primary workloads for our Kepler-47 Series Storage Spaces Direct Clusters are:

  1. Hyper-Converged (Storage Spaces + Hyper-V on each node)
    1. Intel Xeon Processor E3-1270 series with 64GB ECC RAM
    2. Mellanox 25 GbE Dual-Port PCIe for East-West Node-to-Node Communications
    3. 480GB Intel SSD DC S4600 Series Cache Pair to Start
      1. 960GB pair and up for more IOPS
    4. Storage configurable in 12TB, 24TB, 36TB, 42TB, 60TB, 72TB, and soon 84TB in NearLine SAS Drives
      1. Storage Spaces 2-Way Mirror
      2. 1 Capacity Drive’s Space Reserved for Parallel Rebuild
    5. Performance
      1. 25K IOPS @ 70% Read and 30% Write @ 8K
      2. 660MB/Seconds Mean Throughput cache full
  2. Scale-Out File Server (SOFS) on Storage Spaces
    1. Configured as above
    2. Use for Veeam ReFS on Storage Spaces Backup Target
      1. An inexpensive and very scalable solution
    3. Provide a reasonably priced storage target for Hyper-V clusters
      1. Host VHDX files on SOFS

Our Kepler-47 platform can be configured with two or more nodes. At the two node level no switch is required for node to node traffic used for storage I/O. The Mellanox network adapters are direct connected with RoCE RDMA enabled for ultra-low latency and fast access between nodes.

A pair of switches can be added, either the Mellanox MSX1012X 10GbE or the Mellanox SN2010 100GbE/25GbE combination switch, to add more nodes at a later date. We can provide a technology vision for our solutions upon request.

Please e-mail or call for further information!