Exam 3V0-21.21 topic 1 question 34 discussion

B and C?

B & C Which two assumptions could the architect make based on the information from the customer to help size the solution? A. The number of vSphere hosts in a cluster - this won't help size the solution D. The guest operating system for each virtual machine - this won't help size the solution E. The size (CPU/RAM/storage) of the vSphere hosts - this won't help size the solution

A & E are not assumptions, they are what you are sizing in your design D is not needed to size the solution

A and E are the results of the solution and need to be evaluated against the assumptions

we are asked to size the environment, in order to do so we need to know VMs utilization and average size...B,C

B and C : A. The number of vSphere hosts in a cluster -> this is what we want to design B. The average resource utilization of a virtual machine -> This must be assumption in order to design the solution C. The size (CPU/RAM/storage) of the average virtual machine -> This must be assumption in order to design the solution D. The guest operating system for each virtual machine -> who cares ? E. The size (CPU/RAM/storage) of the vSphere hosts -> -> this is what we want to design

From the course book ---Page 19 Analyzing the Current State Use monitoring tools to do an inventory of the existing infrastructure and report the resource usage. For each system, capture peak and average utilization for the following items: • CPU • RAM • IOPS • Network utilization ---Page 96 Calculating the Total CPU and Memory Requirements Multiply the adjusted total number of VMs by the average resource usage per server by using the CPU and RAM usage data from the capacity analysis report at peak utilization times. When sizing hosts, use peak utilization levels rather than average utilization. In this way, all systems can run at their observed peak resource levels simultaneously. ---Page 192 Calculating Datastore Size Average size of all VMDKs for a VM: — Number of VMs x Average size of VM’s VMDK files = VM capacity Sizing for average utilization always introduces the risk of workloads peaking at the same time. During peak utilization, the environment might be constrained for resources.

To provide a solution size of # of hosts for 5000 vms *Assume the vm profile * Calculate resources needed for 5000 vms * Calculate host size, cluster size for the solution Customer can be presented solutions or what if scenarios for different vm profiles.

BC It's SDDC Design. If you go to VMware Cloud Sizer (Advance Sizing) - https://vmc.vmware.com/sizer/advanced-sizing It ask about resource utilization of a virtual machine (In Additional setting) That's Choice B And also it's ask about avg CPU/MEM/Storage of VM (That is Option C)

changed my answer to B & C, see my explanation below.

A & E - As we have no information about the VMs. But we have the requirement the prod environment should be delivered across two data centers (stretched cluster) and we have a service availability target

I don`t think it`s possible to make assumption on vms here since there`s no info provided regarding that. I`d stick with AE since they`re both host related and more feasable to guess.

Only can do it for the hosts

I think A and E. Here is my logic: A. We can initially assume hosts per cluster value. B. This value cannot be assumed because it will dictate the final SDDC solution sizing. (even a minor difference in VM sizing can add up to a big amount when we scale the solution to 5000 VMs) 3. Same as above. This is the key value in solution designing and cannot be assumed. 4. Doesn't impact the solution. 4. This can be assumed initially.

I think A cannot be an assumption because it affects the requirement for availability SLA. E cannot be an assumption either, because it is based on A and can be directly calculated based on assumed size of average VM.

BC? ...This one is hard. I am still not fully convinced for BC over AB, but my logic is - only atribute we could count from customer is VM amount. Without assuming the average resource and size of VM we can't be sure with size of hosts. We know projected availability target but again without vm resource profile can't reliably get amount of hosts. And moreover, AE looks like the solution itself, if we assume that, then where is the solution? But I could be totaly wrong.

I think A and E. The architect can only consider the number and size of ESXi. Virtual machine information is heard from customers and is not considered by architects.