Understand Microsoft Hyper Converged Solution

This whitepaper is written by Romain Serre and Charbel Nemnom which describes Microsoft Hyper-Converged solution in Windows Server 2016 using Storage Spaces Direct, Hyper-V and network technologies. The second part of this document shows an example of this implementation.

 
 
 
 
 
4.9 Star
(15)
18,615 times
Add to favorites
Servers
6/19/2019
E-mail Twitter del.icio.us Digg Facebook
Sign in to ask a question


  • Any 2019 Hyper converged solution material out there ?
    2 Posts | Last post July 01, 2019
    • This is by far one of the best documents on Hyper Converged solution 2016. Have you prepared anything yet on Windows Server 2019? I will be building a new Hyper Converged solution with Windows Server 2019 in two weeks and think it would be really great to refer to a good case study from experts in the field like yourself.
    • I didn't take time yet. Yes I plan to update it. Maybe at the end of the year :) You can also read my blog: https://www.tech-coffee.net. Feel free to contact me if you need help.
  • Link not working
    2 Posts | Last post June 19, 2019
    • Can't download pdf, the link is not working. Can you please upload?
    • Hello,
      
      It should work now.
  • copy and paste
    1 Posts | Last post October 10, 2017
    • bit confused why you would prevent copy and pasting from the document.  typing out powershell is not fun!
  • Can't print the document due to security
    2 Posts | Last post May 16, 2017
    • is there any way to obtain a copy of this document that can be printed? It apparently has security on it that prevents printing.  Thanks in advance,
      
      Brian W.
    • Hi,
      
      Contact me at serre.rom [at] outlook fr. I send you the document.
      
      Regards,
      
      Romain
  • Solution with direct connect network
    4 Posts | Last post April 27, 2016
    • Can the hyper converged solution work with directly connected rdma cards instead of using a switch? I'm considering using mellanox 100gbe cards but a switch is over $10,000. It's cheaper to buy 2 dual cards per server and cable them up directly. Would this solution work?
    • Hi,
      
      The solution needs at least four nodes. So you can't interconnect all nodes in a same subnet (or two) by using direct attached. How would you handle this in the failover cluster configuration ?
      
      For production environment, the switches are a requirement. Even if switches are over 10K$, it is less expensive than standard SAN Fabric. Don't forget that in this kind of solution, the network that's the life :)
    • If you had 3 network interfaces per node, you could connect them as follows: a->b,c,d; b->c,d; c->d; since each connection is bi-directional. Each node could talk to the other 3. Yes?
    • Hi,
      
      It is not a supported scenario: https://blogs.technet.microsoft.com/clausjor/2015/11/23/hardware-options-for-evaluating-storage-spaces-direct-in-technical-preview-4/. In this topic is presented the three network scenario supported.