shares on a VM that is "outside" ANY/all resource pool

Hi i m trying to understand the default behaviour. I have seceral VMs that are NOT in any resource pool.

I am being told that this is a big problem and these VMs will have higher priority for resources (in a contention) higher than the VMs in the default high priority resource pool.

For these VMs I have not set or configured any cistom share or limit settings on the VM - all left at default - on the resources tab they just say - normal and "greyed out" with 1000 in shares area  - also no limit is ticked.

My question is how are these VMs using shares - if i havent set any settings on them - From online im reading the following:

" Ensure that each VM is placed in a resource pool and not out of all resource pools (look at point 4). VMs outside pools with shares configured could get more resources than VMs inside resource pool. "

 - now given I have  NOT touched shares for these VMs that are outside any resource pool - are they "configured for shares" at all? and as they are not is the advice im getting not in fact correct. Do these homeless VMs take priority anyway? thanks  -  Andrew Hancock - still here been great help in the past :)
LVL 1
philb19Asked:
Who is Participating?
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:
I am being told that this is a big problem and these VMs will have higher priority for resources (in a contention) higher than the VMs in the default high priority resource pool.

some truth in the statement, but a big problem.... and it depends what has been allocated, and if the resource pool, runs out of resources.

The resource pool you have defined is a guaranteed ring fence pool of resources for the VMs inside it, the VMs outside this pool, will clearly not be affected by the constraints of the resource pool.

and this only matters, if resources on your host, or cluster are under contention, e.g. heavily loaded/over subscribbed, and these days we don't see many over subscribbed hosts, unless you are on a real low budget, and have the world running on a single host.

e.g. we have Resource Pools setup for Students, we want to ring fence what resources we give the VMs in that resource pool. we do not want their VMs, to affect Production workloads, that are NOT in resource pools.

The wording "big problem" is a strong phrase.
0
philb19Author Commented:
Thanks Andrew. He has in fact deleted all the resource pools now (default ones -  high,medium and low) - so now there are no resource pools - is this the best approach?

Bearing in mind I have simply steared away from customizing resource pools altogether - only mistake made possibley not putting some vms in any of the 3 default resource pools
0
philb19Author Commented:
"with machines not being placed in ANY resource pool meaning that they actually had higher priority and more shares than those that were placed in the high priority pool! -"

do homeless VMs  by default get more shares ? as he states above
0
Powerful Yet Easy-to-Use Network Monitoring

Identify excessive bandwidth utilization or unexpected application traffic with SolarWinds Bandwidth Analyzer Pack.

Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Thanks Andrew. He has in fact deleted all the resource pools now (default ones -  high,medium and low) - so now there are no resource pools - is this the best approach?

Keep it simple, you start using resource pools, you've got to start getting the calculator out working out what VM has what resources!

You ONLY really need resource pools in my opinion, if you are "leasing out resources" to different companies, or a Dept has paid for a certain amount of CPU and Memory.

otherwise, ALL VMs will receive the same priority as all the other VMs.

Unless, you are really really pushed for resources, and then you need to start thinking of which Services need more resources than others!

If there is NO constraint, they all have the same!

VMs not in a resource pool, not affected by the resource pool they are in, so it could be seen they have higher priority, but all things being equal, only when resources are tight!
0
David Johnson, CD, MVPOwnerCommented:
Everything has a finite limit of resources. CPU/Memory/DISK IOPS and by using resource pools you are allocating limitations and prioritization of available resources. The default high/medium/low are simply guidelines and you are free to modify and create another resource pool. It depends upon your business's priorities how you allocate your available resources. If items are not in managed in a resource pool then a group of users could tie up all of your resources leaving business critical applications starving for resources.  A system engineer overprovisions the resources as much as they can within their budget. This works fine under normal workloads until you get a bunch of users start rendering videos, while accounting is trying to push out the payroll.
0
philb19Author Commented:
Hi - thanks Andrew David. Ok then so I take it then that "shares"/resource pools ONLY come into play when a host has contention for resources. The advice I am getting is that there has been some CPU or other resource contention. The issue is I seriously doubt that there has been. The host is not over subscribed.

Now the question is "can I prove this" ? for say the last month - simply using vcenter client performance monitor (will this suffice to prove the point host hasnt been under cpu pressure/contention? ) THe contractor was using xangati -  Xangati is a multi-hypervisor management suite that provides exceptional visibility, predictive and historical analysis and service assurance analytics.

All along ive been getting "no contention" seen from the same contractor (using logic monitor)!! - but given that they are now using what looks like a new pretty comprahensive new  tool - Is it feasble to get back to them with VMware native performance monitor graph for the last month? - thanks
0
philb19Author Commented:
also can i get a native vmware perf mon from vcenter - on the resource pools themselves ?
0
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Yes we use Xangati it has lots of bells and whistles....

What do there metrics show and prove have you seen them.

Most products just grab the data from vCenter Server and interpret it differently.
0
philb19Author Commented:
No havent seen yet been away from the office - They are saying there was some contention.

I dont quite get how logic monitor and native tools wouldnt pick up contention ? Are we saying the native performance monitor is not good enough? If Xangati is just grabbing from vcenter anyway - interperet differently ?- what meaning interpretting correctly? and logic monitor and vcenter report erronous information - Thanks
0
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
There will always be contention....

BUT was performance of a VM impacted.

I'll give you a trial URL that uploads data from vCenter and processes and gives you a report...
0
philb19Author Commented:
ok thats interesting - so there can be contention without VM being impacted ? how so ? If a VM is struggling for resources I would have thought it would have to be impacted (made to wait / run slower - effect application layer)

thanks for trial - will this give me historical perf data?
0
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Contention is subjective and does happen, a VM can be struggling for resources, but it's performance might be impacted or may not be impacted.

stating

"They are saying there was some contention." it's not very helpful, where, when, what VM.... unless the answer is provided by another Bill or Days Consultancy to tell you how to fix it!

Do you have VMware HA and DRS ? you can see if the hypervisor managed to give resources to a VM, with DRS

e.g. you have give a VM 4 vCPU @ 2.GHz, and 16GB RAM, but they may be a smaller period of time, the VM, did not get 8GHz of CPU, and 16GB RAM, e.g. if only had 7.5GHz, and 14GB RAM

BUT was the performance of the VM impacted or struggle....

By nature of server consolidation, and using a hypervisor, you are making a compromise...otherwise you would put your VM, on a physical single server.

You can use VMware vCenter Server Performance charts, Veeam One, VMturbo, VMware Operations Manager, Xangati, Opvizor, Xtravirt SONAR - and they all give you a different result...

So are the VMs performance being impacted....

you are correct, VMs run slower, e.g. SQL reports, Terminal Services etc

But the four metrics, CPU, Memory, Disk and Network are the metrics to check and observe!

this is the trial URL...

http://www.opvizor.com/
0

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
philb19Author Commented:
Thnaks Andrew + David - Awesome help as usual. - Ill be back with another question in a tick :)
0
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.