Azure HCI VM Create Capacity

Mehrdad Ghasemi 1 Reputation point
2021-02-17T08:56:58.677+00:00

Hello
In HCI structures, for example, if we have 2 servers that have 64 GB of RAM on each, that is, there is a total of 128 GB of RAM.
Can a maximum of 64 GB of RAM be allocated to build a VM?
Can I allocate 100 GB of RAM to a VM?
Thanks

Azure Stack HCI
Azure Stack HCI
A hyperconverged infrastructure operating system delivered as an Azure service that provides security, performance, and feature updates.
358 questions
0 comments No comments
{count} votes

2 answers

Sort by: Most helpful
  1. CyrAz 5,181 Reputation points
    2021-02-17T09:34:29.42+00:00

    No, you can't allocate more compute (RAM/CPU) resource than what's available on a single host.

    0 comments No comments

  2. MattMcSpirit-MSFT 561 Reputation points
    2021-02-18T23:22:12.64+00:00

    Well, yes and no. From a CPU perspective, Hyper-V does a great job of managing CPU scheduling, and while you can't create a VM with more vCPUs than you have logical processors in your host, the total number of vCPUs allocated across VMs can add up to more than the total logical processors in the host. Hyper-V will use it's advanced scheduling to manage this. If all your VMs spike to 100% CPU at once though, there's only so much actual processor capacity that Hyper-V can allocate.

    Good write up here: https://www.altaro.com/hyper-v/hyper-v-virtual-cpus-explained/

    From a memory perspective, with Dynamic Memory, you can set a VM to have a 1TB memory if you like, yet again, Hyper-V will manage what it actually allocates to the VMs, so in your case, the VM will never be allocated more than 64GB minus what the host needs itself. If you have a number of VMs running, and all have 1TB defined as their maximum memory, Hyper-V will dynamically assign memory to and from VMs based on their demand.

    Good write up here: https://www.altaro.com/hyper-v/hyper-v-dynamic-memory-explanation-and-recommendations-2/

    0 comments No comments

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.