Converged Storage - Some Available Options - Part 2

This is the second part of a two part article on converged storage solutions.  Part 1 is here.  Part 1 dealt with vSphere VSANEMC ScaleIO,​ Atlantis ILIO USXStorMagic SvSAN and Maxta SDS.  In this second instalment I'll try to consisely summarise:

​​Just to mention again something I said in part 1; not all of these solutions are 'merely' VSAs, some represent complete hardware/software 'hyper-converged' storage solutions.  The simple aim of this article is to group these similar technologies together.

Parallels Cloud Storage

  • vSAN with replication and DR functionality
  • Sits on commodity hardware
  • Runs as part of Parallels overall virtualisation solution (Parallels Cloud Server product)
  • Management is via Parallels management toolset
  • SSD caching is optional
  • Scale-out to petabytes of storage
  • main components are MSD (Meta Data Server) and CS (Chunk Server)

Nexenta Connect

  • Complete VDI solution. Not just a converged storage solution.  E.g. desktop management software included
  • Provides a virtual SAN to all hosts within an ESX cluster
  • VSA provides local data store abstraction and offers it out as NFS export for consumption by vSphere ESX hosts
  • Not limited to just DAS.  Can make use of SAN storage too
  • Data integrity improvements for DAS through use of ZFS as underlying file system
  • SSDs can be used for caching purposes, enhancing overall performance
  • Separate management VMs reside on specific host in cluster
  • Advanced toolset for benchmarking and performance optimisation
  • ​Software only solution allowing use of commodity hardware (on VMWare HCL, of course)


  • Hardware/software 'hyper-converged solution.  No included custom hardware (h/w is standard x86, rebadged) 
  • Appliance VM (CVM) provides the storage abstraction (nodes create clustered FS, export as NFS)
  • File system export from appliance (NFS) is consumed by vSphere as a datastore
  • Hardware platform contains SSD and HDD storage
  • Distributed file system (NDFS) abstracts storage complexities, presents centralised array
  • Hypervisor options are VMWare ESX, MS Hyper-V and KVM
  • ​Advanced performance enhancing features such as 'shadow clones', inline data deduplication (cluster-wide)
  • Supports VAAI (VMWare API for Array Integration), MS ODX (Offloaded Data Transfers)
  • Disk balancing to give uniform data distribution across cluster

​Simplivity OmniCube

  • In the news a lot recently with rumoured buyout by HP (see here)
  • Hardware/software hyper-converged storage solution
  • Simple box scale-out architecture with VM-centric single point of management
  • Value adds such as remote data replication and public cloud integration (AWS VPC)
  • Inline deduplication and compression via custom PCIe card
  • Single hypervisor option - VMWare vSphere
  • Compatible with standard VMWare tools such as vMotion, Storage vMotion, HA DRS
  • Several layers of cache provide enhanced performance
  • Omnicube Virtual Controller VM is key component
  • Omnicube Accelerator Card OAC is key hardware component
  • Software architecture arguably fairly similar to Nutanix
  • ​NFS is exported file system for consumption (as NFS datastore) by VMWare vSphere