5V0-22-23 Sample Questions & Answers

For maximum space efficiency while tolerating a single failure (FTT=1), RAID-5 erasure coding is the best choice. It requires a minimum of 4 hosts. RAID-1 (Mirroring) provides FTT=1 but consumes twice the storage space. RAID-6 provides FTT=2, which is more protection than required and is less space-efficient than RAID-5. Given the priority is space efficiency with single-host failure protection on an 8-node cluster, RAID-5 is the optimal configuration.

The 'What if analysis' in the vSAN capacity monitor is a planning tool that calculates usable capacity based on a selected storage policy. To provide a conservative, worst-case estimate, it assumes no space savings from deduplication and compression (i.e., a 1.0x ratio). The actual usable capacity will be higher if the data achieves a better-than-1.0x ratio.

'Full data migration' is the correct option for permanently decommissioning a host. This mode evacuates all vSAN components from the host's disks to other hosts in the cluster, ensuring that all objects remain compliant with their storage policies. Once the migration is complete, the host can be safely removed. 'Ensure accessibility' only moves the necessary data to maintain availability and is intended for temporary maintenance. 'No data migration' is used for very short-term work where a temporary loss of redundancy is acceptable.

In vsantop, backend congestion is a measure of how much I/O is being queued due to the underlying storage devices' inability to service requests quickly enough. A high value indicates that the physical disks are the bottleneck. This could be due to an under-provisioned cache tier, slow capacity devices, or a failing device that is introducing high latency.

For a vSAN Stretched Cluster, the network latency requirements are strict. The RTT between the two main data sites must be 5ms or less for synchronous I/O replication. The RTT between each data site and the witness site can be much higher, up to 200ms, because only metadata and witness component updates are sent to the witness, not the full VM I/O.

While in newer versions of vSphere this is enabled by default and managed through the UI, the specific command-line method to enable or disable TRIM/UNMAP on a vSAN cluster is esxcli vsan cluster set --unmap=true. This command needs to be run on one of the hosts in the cluster.

Deduplication and compression are powerful space-saving features in vSAN OSA, but they require the performance of an All-Flash configuration. These features are not available on hybrid vSAN clusters. For VDI workloads, which often have highly redundant data across desktops, the space savings from deduplication can be substantial, potentially offsetting the higher cost of flash storage.

In an HCI Mesh topology, the cluster that shares its datastore is the 'server' cluster, and the cluster that consumes the remote datastore is the 'client' cluster. To use the high-performance storage from Cluster-A for a workload running on Cluster-B, Cluster-A must be configured as the server, and Cluster-B must be the client that mounts the remote vSAN datastore from Cluster-A.

Storage vMotion is the standard vSphere feature designed for migrating a running virtual machine's files from one datastore to another without service interruption. Since the vSAN datastore and the FC SAN datastore are both accessible by the ESXi hosts, Storage vMotion is the most direct and efficient method to accomplish this task and free up space on the vSAN cluster.

The Cluster Level Object Manager (CLOM) is responsible for all object-related operations in the vSAN cluster. Its primary role is to receive storage policy creation/modification requests, determine the necessary component layout to satisfy the policy (e.g., RAID level, FTT), and instruct the Distributed Object Manager (DOM) to create and place those components on hosts across the cluster.