An architect is designing a solution for an environment with two types of resource profiles that must be virtualized. The first type consists of Tier 1 virtual machines that are disk I/O intensive, but do NOT require high CPU or memory. The second type consists of Tier 2 virtual machines that require a lower CPU and memory allocation and have minimal disk I/O.
Which design recommendation should the architect make for distributing the resource profiles?
When designing a solution for two types of resource profiles with different disk I/O and CPU/memory requirements, it's optimal to separate them into different clusters. This allows for better resource allocation and management. The Tier 1 cluster, which is disk I/O intensive, should have fast storage. The Tier 2 cluster, requiring lower CPU and memory allocation with minimal disk I/O, does not need high-performance storage. This separation ensures that each resource profile gets the most suitable hardware configuration, enhancing performance and efficiency.
Changing my answer to B. Both resource profiles DO NOT require high CPU or memory, only a different level of disk IOPS. We can use vSAN here to apply a storage policy to the the different tier VM's
good observation! that's the only difference
Ok but no one talks about usage of VSAN here. Technically both Tiers could run on same cluster in an average IT company. But if we focus what is desired in ideal world, separating tiers to different clusters would be the best approach in my opinion. So that`s why I`m still with "A".
Answer is A. Checked in the training manual, chapter 10 Infrastructure Manageability: When mention about tiers... always talks about DIFFERENT CLUSTERS. Vmware suggest this in order to calculate spare hosts, cpu contention, disk performance, etc. In this case B and D should be disregarded, you should not mix Tier 1 and 2 in the same cluster. A option meets requirements: faster storage
I agree. Just checked the same doc and I agree with your comment
I think the key to the answer is the word "profile", you can have different profiles in the same hardware. So, the answer is B.
Might be D. Just create two resource pools and assign shares within pool.
Agree with this.
i think is A
The answer is A
IOPS performance profile for each app tier are created in the storage and vSphere storage policies. There is no requirement to say sharing the host hardware or array will affect storage performance. Therefore without more NFR option B.
Usually you'd have one cluster and a storage device with differently performing storage tiers. I suppose it's A for this specific question. I also suppose that 75% of the test takers will choose the wrong answer here.
Storage array has storage tiering function that allow classification of of iops workload
The same storage can handle high IOPS and low IOPS so it is not question to have to sperate storage for each workload
B = Same profile and same hardware. Did not mention about storage. C = More disk space but doesnt talk about IOPS. So requirement is not met. D = Fast CPU / Memory is not required for both Tier of VMs. Hence it is a expensive solution. Hence answer is A. It fulfills the requirement and follows the best design practice of isolating VM Tiers.
"A" looks like the best approach if there`s no budget constraints. But in a real life scenario, we all know that, "B" is what is being applied in most companies. I`m not sure from which perspective we should evaluate this question. I hope this question doesn`t appear in the exam as there`re two correct answers in my understanding. Please provide more concrete info on this one if you have.
Agree with Mariosoftnet. Doubting between A and B, so following best practices as there is no cost contraints.
B Fast CPU and large memory aren't required.
My answer is D, Regardless of fast CPU and Memory. The requirement is the fastest storage, which is only mentioned in D.
We DON'T need host hardware that has fast CPU, large amounts of memory. That's the key difference between B and D.
D seems to be correct, as the some VMs are disk intensive while others are not. so we need fast disk storage with resource profiles.