Esxi host with mixed cpu cores.

Have an infrastructure  question , hope some one will suggest a good solution.
Client has 6 esxi hosts in a cluster.
Two of the servers are with 4 core dual CPU.
Three of them are with 8 core dual cpu.
One of them is with 12 core dual cpu and the memories are not equal in the esxi hosts too,  this case the DRS is unbalanced status.
It is time for them to get replaced the old servers with new server possibly upgrade the vsphere to ver6.0
Definitely the three servers with 4 core require immediate attention (replace).
i suggested my client to replace with 8 core dual cpu so the cluster will be in balance, that is, all the servers will have almost same cpu and memory.
The system admin prefers to get latest server 12 or 18 core, this case the cluster will have three 18 core, two 8 core and one 12 core.
And He said that he can set affinity rule so his tier -1 servers will be staying on the 18 core esxi hosts and he does not need to worry about failure of a single host, that is, the cluster will still have two more 18core servers to fail over.
i suggested him that if you are not going to have a vm with 32vcpu or more why you bother you want to spend money on 18 core instead stay with 8 core.
What you will suggest if you are in my situation?
LVL 2
sara2000Asked:
Who is Participating?

[Product update] Infrastructure Analysis Tool is now available with Business Accounts.Learn More

x
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Remember DRS IS NOT A LOAD BALANCER!

e.g. EVEN Clusters with the same CPU Models and Cores, and Identical Memory in ALL Hosts, you may see in your words unbalanced STATUS!

Do not expect to see 50% CPU, 50% Memory, on ALL HOSTS!

If you think it is you are WRONG!

DRS is designed to ensure, that Virtual Machines, obtain the correct CPU and Memory when they ask for the resources.

So have you actually checked to see if Virtual Machines, are NOT being DELIVERED the resources they are requesting!

So 4 Cores may be old hat, but I would be more concerned of which maches are NOT on the HCL, or/and which machines are out of a Support Contract.

I'm not sure I would much around with setting affinities.

Memory is often the bottle neck, and more important than cores.

In an ideal world, we would love all hosts to be equal, same processors, makes, models, cores, same memory, but we know what does not happen!

Also make sure, that HA also works, should a server fail!

If you have super duper, 12, 18 core servers, with the memory, if this server should fail, you may not have enough "slots" left in the cluster to take the load.....
sara2000Author Commented:
Andrew, Thank you for your quick reply.
If i understood correctly, DRS is confusing me lot . when you say "Remember DRS IS NOT A LOAD BALANCER!"
This means that VM will not be migrated to another esxi host if that vm demands more resources form the hosting esxi which do not have to spare, only DRS will be useful when you restart the vm the vm will be housed in the appropriate esxi host which has enough resources , am I correct ?

Having 18 core in the cluster is going to introduce slot issue this case better stay with same core and memory , Thant what you would suggest?
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
DRS is complex. again never expect ALL your hosts to be equal with CPU and Memory, you will be waiting a long time!

Basically, if the host cannot deliver the required CPU and Memory resources entitlement, the VM asks for at startup or during operation, it will be Migrated or started on the correct host which can deliver those resources.

This also depends on the setting of DRS, and it's best to have set to Fully Automated, - Aggressive!

Again, equal hosts make it easier, CPU Cores and Memory.

As an aside, if you do not have the following volumes on your bookshelf, I would *HIGHLY RECOMMEND* them. These are the best source of VMware vSphere HA and DRS in the world!

VMware vSphere HA and DRS Technical deepdive

By Duncan Epping and Frank Denneman

(http://www.yellow-bricks.com/vmware-high-availability-deepdiv/)

Written by Duncan Epping and Frank Denneman, both of whom are Consulting Architects at VMware and are perceived by the industry as Subject Matter experts on VMware High Availability and VMware Distributed Resource Scheduler.

This book zooms in on two key components of every VMware based infrastructure. It covers the basic steps needed to create a VMware HA and DRS cluster, and goes on to explain the concepts and mechanisms behind HA and DRS which will enable you to make well educated decisions. You will get the tools to understand and implement e.g. HA admission control policies, DRS resource pools and resource allocation settings and more.

VMware vSphere 5.1 Clustering Deepdive on Amazon
sara2000Author Commented:
Thank you for this,  the reason I want to have three 18 core esxi host  is that the application vendors might require 32 core or more  cpu etc soon. This case my old esxi host (4 core)  is not going to provide that requrement  (dual 8 core cpu hyper -threads =32). Instead place three 18 core esxi host and set affinity rule to place the application servers in that higher core (18 core) esxi host.
Set the DRS to 20% failure than one esxi host failure, here i do not need to worry about the slot size?.
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Thank you for this,  the reason I want to have three 18 core esxi host  is that the application vendors might require 32 core or more  cpu etc soon. This case my old esxi host (4 core)  is not going to provide that requrement  (dual 8 core cpu hyper -threads =32). Instead place three 18 core esxi host and set affinity rule to place the application servers in that higher core (18 core) esxi host.
Set the DRS to 20% failure than one esxi host failure, here i do not need to worry about the slot size?.

Fair enough and valid reasons to move forward, BUT just be aware of

1. HA failure, and enough resources left for a server failure.
2. Do not expect to have equal loaded hosts!
3. Look at capacity, graphs, cluster resources,  and plan!

and if you don't have vCenter Operations Manager - get it or vCOPs,. or vROPS

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
VMware

From novice to tech pro — start learning today.