Resource Management and Availability
Free VMware Certified Professional - Data Center Virtualization (vSphere 8) practice — 6 questions on Resource Management and Availability, with explanations. No sign-up. Full 12-question mixed test →
You have a resource pool named 'Finance-RP' configured with a Memory reservation of 16 GB and Expandable Reservation enabled. VMs already running inside the pool consume the full 16 GB in reservations. An administrator attempts to power on an additional VM configured with a 4 GB memory reservation. The parent resource pool currently has 32 GB of unreserved memory available. What happens when the VM is powered on?
When Expandable Reservation is enabled on a resource pool, and the pool's own configured reservation is fully committed, the pool can dynamically borrow unused reservation capacity from its parent (up the hierarchy) as long as the parent has enough unreserved capacity — allowing the VM to power on.
A vSphere HA cluster contains 8 identical ESXi hosts, each with 128 GB of physical memory. The cluster's admission control policy is set to 'Cluster resource percentage' with both CPU and memory failover capacity set to 25%. Ignoring virtualization overhead, what is the maximum amount of memory available for powered-on VM reservations in this cluster?
Total cluster memory is 8 x 128 GB = 1024 GB. With a percentage-based policy reserving 25% for failover, 256 GB is set aside, leaving 1024 - 256 = 768 GB available for VM reservations.
An administrator creates a VM/Host affinity rule of type 'Must run on hosts in group' that binds a group of database VMs to a host group containing only Host-A and Host-B. DRS is disabled in this cluster. Host-A fails unexpectedly, and Host-B currently has no free resources available for the affected VMs. vSphere HA attempts to restart the failed VMs. What is the outcome?
'Must run on' VM/Host affinity rules are mandatory constraints enforced by both DRS and HA and can never be violated, even during an HA failover. If no host within the allowed group has sufficient capacity, the VM restart fails and the VM remains unpowered until capacity becomes available within the group.
In vSphere 8, what is the maximum number of vCPUs supported for a single virtual machine protected by vSphere Fault Tolerance?
vSphere Fault Tolerance supports FT-protected VMs with up to 8 vCPUs, a limit increased from the earlier 4-vCPU maximum in older vSphere releases, and this remains the supported maximum in vSphere 8.
What is the default 'APD timeout' value in vSphere HA before an ESXi host formally declares a datastore to be in an All Paths Down (APD) state, potentially triggering the configured HA APD response?
vSphere uses a fixed default APD timeout of 140 seconds. Once this timeout elapses without I/O recovery, the host formally declares the device as APD, which can then trigger the cluster's configured 'Response for Datastore with APD' HA setting.
A cluster contains two top-level resource pools: 'RP-Gold' configured with High CPU shares, and 'RP-Silver' configured with Normal CPU shares. Inside RP-Silver, a VM named 'Web-01' is individually configured with Custom CPU shares set to a very high numeric value (10,000). During CPU contention, how does DRS treat Web-01 relative to VMs residing directly inside RP-Gold?
This is the resource pool 'sibling rivalry' behavior: shares are only meaningful relative to sibling objects sharing the same parent in the resource hierarchy. Web-01's high custom share value only influences its priority among other VMs directly inside RP-Silver. When RP-Silver as a whole competes against RP-Gold for the parent's resources, only RP-Silver's own pool-level share setting (Normal) is used in that comparison.
Ready for the real thing?
The full course has two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed answer explanations.
Start my full course on Udemy →