Solved

virtualisation - physical disks and the best way to order them

Posted on 2016-08-25
10
77 Views
Last Modified: 2016-08-28
Hi Everyone,

Building a VM environment tomorrow and want to know the best use of what I have.
I did get some advice on another question but I actually really did take it off topic, so new question to be fair.

I have 12x 300GB physical disks.

I want to have the following roles:

CRM+SQL
Exchange
File
Print
DHCP
DNS
DC
RDP (For a few remote users)

How would everyone set this up as all of these roles with be VMs in Windows Server 2012 R2.
I have ran a fair few scenarios... but I either am left with not enough space for File services ( this is an existing site and I need 900GB before pruning - no time to prune now :) ) or several VMs on the same one massive VIRTUAL DISK made by the RAID controller.

Can I have your thoughts please.

Many thanks in advance

R
0
Comment
Question by:roycasella
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
  • 5
  • 2
  • 2
  • +1
10 Comments
 
LVL 120
ID: 41770760
You need to have a PAIR of DISKS for the Hypervisor OS - e.g. System C: Drive

and then RAID 10 or RAID 6 the remaining disks for DATA (VM Storage) for ALL your VMs, as VHD/VHDX.

see this EE Article, written by our very own Hyper-V Expert - Philip Elder

Some Hyper-V Hardware and Software Best Practices
0
 
LVL 7

Author Comment

by:roycasella
ID: 41770777
Cool Thanks Andrew.

Read through... nice read.

So it fits in with what I was thinking of:

2x disks RAID1: HOST OS

10x disks either RAID6 or RAID10 (RAID 6 I think for storage: VM Hard Drives, data etc.... All all of the VMs for all of the roles will sit on this one large RAID ARRAY of 10x disks.

Andrew, would you create one single volume for the VMs, or would you chop it up into separate volumes in Windows Server 2012 R2.. i.e. Volume for CRM + SQL, Volume for Exchange, Volume for Print etc...

Many thanks in advance

R
0
 
LVL 120

Accepted Solution

by:
Andrew Hancock (VMware vExpert / EE MVE^2) earned 250 total points
ID: 41770818
I would use ALL the disks, with a single volume.
0
Space-Age Communications Transitions to DevOps

ViaSat, a global provider of satellite and wireless communications, securely connects businesses, governments, and organizations to the Internet. Learn how ViaSat’s Network Solutions Engineer, drove the transition from a traditional network support to a DevOps-centric model.

 
LVL 7

Author Comment

by:roycasella
ID: 41770851
OK Excellent... I will RAID 1 - 2x disks for Host OS and then  RAID 6 and have one big volume..  for all VMs... that works great for me. Just wanted to check I was not making any silly performance NO NOs putting on one volume.

Thanks again Andrew

R
0
 
LVL 39

Assisted Solution

by:Philip Elder
Philip Elder earned 250 total points
ID: 41770867
I have an EE article that may help with your questions: Some Hyper-V Hardware and Software Best Practices.

One RAID 6 array with two logical disks:
75GB for host OS
Balance for VM settings and VHDX files.

This allows us to spot rebuild the host if something kills it with ease. We leave a bootable USB flash drive plugged in to the server and have an iDRAC Enterprise/iLO Advanced/Intel RMM to allow us to boot and re-install/reconfigure the hosts. Import the existing guests and we're back in business. BTDT

Also, make sure the RAID controller has at least 1GB of non-volatile cache memory.
0
 
LVL 7

Author Comment

by:roycasella
ID: 41772354
Thanks guys.

Yes RAID controller all good.

So Phillip .. just to confirm as I am doing it tomorrow (:))... You are suggesting that I RAID 6 the entire array of 12x disks and create two logical volumes. One for OS and one for Data being the VMs and all their data etc...

As opposed to RAID 1 2x drives for HOST OS and Raid 6 10x for Data? This is what I was thinking but wanting some steer on. I thought that would perform better the two arrays.. It might be a dumb statement

Many thanks in advance

R
0
 
LVL 55

Expert Comment

by:andyalder
ID: 41772386
The advantage of having the OS on separate disks is that if the OS/hypervisor goes wrong you can just pull the data disks out and import them into another server and data recovery can be easier as software de-striping tools work easier with a single logical disk on an array.

The advantage of having all the disks in one group and then two logical disks on it is that once the host OS is booted it's pretty idle so those two disks add more available IOPS to the VMs.

For SQL full recovery there's an advantage of having the transaction logs on different spindles than the data, then if the data array fails the backup plus transaction logs bring you right up to date rather than losing what you did since last backup, the two array model lets you put the logs on those host/hypervisor disks in a separate partition.
0
 
LVL 7

Author Comment

by:roycasella
ID: 41772393
Thanks Andy.. great point.

I would personally put in RAID 10 if I could afford the disk capacity lost and get some great gains.. But RAID 6 looks like it needs to be.

Nice suggestions from this thread.. Thanks everyone

R
0
 
LVL 39

Expert Comment

by:Philip Elder
ID: 41773371
That is correct. Full IOPS are available to the virtualization setup.
0
 
LVL 7

Author Closing Comment

by:roycasella
ID: 41773699
Gone ahead with this setup. Happy with that.

Thanks everyone
0

Featured Post

Problems using Powershell and Active Directory?

Managing Active Directory does not always have to be complicated.  If you are spending more time trying instead of doing, then it's time to look at something else. For nearly 20 years, AD admins around the world have used one tool for day-to-day AD management: Hyena. Discover why

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

The following article is comprised of the pearls we have garnered deploying virtualization solutions since Virtual Server 2005 and subsequent 2008 RTM+ Hyper-V in standalone and clustered environments.
In this article, I will show you HOW TO: Perform a Physical to Virtual (P2V) Conversion the easy way from a computer backup (image).
In this Micro Tutorial viewers will learn how to restore single file or folder from Bare Metal backup image of their system. Tutorial shows how to restore files and folders from system backup. Often it is not needed to restore entire system when onl…
This tutorial will walk an individual through the process of configuring their Windows Server 2012 domain controller to synchronize its time with a trusted, external resource. Use Google, Bing, or other preferred search engine to locate trusted NTP …

738 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question