Storage interview questions

1. A.mapping a physical LUN directly to a VM
2. B.ESXi's pluggable storage architecture that manages multiple physical paths to a LUN for redundancy and load balancing (Round Robin, Fixed, MRU)
3. C.a point-in-time delta state of a VM — not a backup
4. D.pre-allocating and zeroing a virtual disk upfront

Answer + AI explanation with Pro

1. A.VVols (Virtual Volumes)
2. B.snapshot
3. C.Datastore
4. D.RDM

Answer + AI explanation with Pro

1. A.snapshot — NVMe over Fabrics (RDMA, TCP, or FC) gives ESXi low-latency block storage with deeper queues than SCSI.
2. B.snapshot — a point-in-time delta state of a VM — not a backup
3. C.snapshot — Allocating more virtual disk capacity than physically exists, relying on actual usage staying below the total; requires monitoring to avoid datastore-full outages
4. D.snapshot — a clone sharing the base disk with its parent to save space

Answer + AI explanation with Pro

1. A.a clone sharing the base disk with its parent to save space
2. B.VMware HCI software-defined storage pooling local disks
3. C.A mapping file on VMFS that gives a VM direct block access to a SAN LUN, used for clustering software or storage-array tools
4. D.A thick disk whose blocks are allocated up front but zeroed only on first write, giving fast creation with a small first-write penalty

Answer + AI explanation with Pro

1. A.NVMe over Fabrics (NVMe-oF)
2. B.Datastore cluster
3. C.linked clone
4. D.VVols (Virtual Volumes)

Answer + AI explanation with Pro

1. A.linked clone — a clone sharing the base disk with its parent to save space
2. B.linked clone — A datastore-wide feature that enforces fair IOPS distribution by share value when latency crosses a congestion threshold
3. C.linked clone — Balances VMs across a datastore cluster by space usage and I/O latency, with affinity and anti-affinity rules.
4. D.linked clone — Allocating more virtual disk capacity than physically exists, relying on actual usage staying below the total; requires monitoring to avoid datastore-full outages

Answer + AI explanation with Pro

1. A.Hardware-offload operations (ATS, XCOPY clone, Zero, UNMAP) that move work from the host to the array.
2. B.A group of datastores managed as one capacity pool, the unit Storage DRS load-balances across.
3. C.A thick disk whose blocks are allocated up front but zeroed only on first write, giving fast creation with a small first-write penalty
4. D.an independent complete copy of a VM

Answer + AI explanation with Pro

1. A.Raw Device Mapping (RDM)
2. B.NVMe-oF
3. C.Datastore cluster with Storage DRS
4. D.full clone

Answer + AI explanation with Pro

1. A.full clone — an independent complete copy of a VM
2. B.full clone — NVMe over Fabrics (RDMA, TCP, or FC) gives ESXi low-latency block storage with deeper queues than SCSI.
3. C.full clone — A mapping file on VMFS that gives a VM direct block access to a SAN LUN, used for clustering software or storage-array tools
4. D.full clone — VMware HCI software-defined storage pooling local disks

Answer + AI explanation with Pro

1. A.Tracks disk blocks modified since the last backup so incrementals copy only changes.
2. B.Balances VMs across a datastore cluster by space usage and I/O latency, with affinity and anti-affinity rules.
3. C.Returns freed thin-provisioned blocks to the array; ESXi issues automatic UNMAP on VMFS-6.
4. D.allocating disk space on demand rather than upfront

Answer + AI explanation with Pro

1. A.RDM
2. B.vVols and SPBM
3. C.Lazy-zeroed thick disk
4. D.thin provisioning

Answer + AI explanation with Pro

1. A.thin provisioning — pre-allocating and zeroing a virtual disk upfront
2. B.thin provisioning — allocating disk space on demand rather than upfront
3. C.thin provisioning — A transport (over RDMA, TCP or Fibre Channel) that extends the low-latency NVMe protocol to networked storage, supported as a datastore type in vSphere 8
4. D.thin provisioning — vStorage APIs for Array Integration: offloads operations like full copy, block zeroing and hardware-assisted locking (ATS) to the storage array

Answer + AI explanation with Pro

1. A.Telling the array which blocks a thin disk no longer uses so freed space returns to the pool; automatic on VMFS-6 at a configurable rate
2. B.allocating disk space on demand rather than upfront
3. C.pre-allocating and zeroing a virtual disk upfront
4. D.A transport (over RDMA, TCP or Fibre Channel) that extends the low-latency NVMe protocol to networked storage, supported as a datastore type in vSphere 8

Answer + AI explanation with Pro

1. A.Storage Policy-Based Management (SPBM)
2. B.VAAI primitives
3. C.Pluggable Storage Architecture (PSA)
4. D.thick eager-zeroed

Answer + AI explanation with Pro

1. A.thick eager-zeroed — A framework where each virtual disk is a native array object managed via VASA, enabling per-VM array snapshots, replication and policy-based placement
2. B.thick eager-zeroed — Independent-persistent disks are unaffected by snapshots while independent-nonpersistent disks discard all writes at power-off
3. C.thick eager-zeroed — pre-allocating and zeroing a virtual disk upfront
4. D.thick eager-zeroed — an independent complete copy of a VM

Answer + AI explanation with Pro

1. A.the KVM copy-on-write disk format supporting snapshots
2. B.VMware's clustered file system for block storage
3. C.Distributes I/O across all active paths; the IOPS-per-path limit (default 1000) is often tuned to 1 for all-flash arrays.
4. D.Independent-persistent disks are unaffected by snapshots while independent-nonpersistent disks discard all writes at power-off

Answer + AI explanation with Pro

1. A.VMFS
2. B.Thin provisioning oversubscription
3. C.snapshot
4. D.Storage I/O Control (SIOC)

Answer + AI explanation with Pro

1. A.VMFS — A datastore-wide feature that enforces fair IOPS distribution by share value when latency crosses a congestion threshold
2. B.VMFS — A group of datastores balanced by Storage DRS, which migrates VMDKs based on space usage and I/O latency and enforces VMDK anti-affinity
3. C.VMFS — VMware's clustered file system for block storage
4. D.VMFS — Independent-persistent disks are unaffected by snapshots while independent-nonpersistent disks discard all writes at power-off

Answer + AI explanation with Pro

1. A.the KVM copy-on-write disk format supporting snapshots
2. B.A group of datastores managed as one capacity pool, the unit Storage DRS load-balances across.
3. C.mapping a physical LUN directly to a VM
4. D.pre-allocating and zeroing a virtual disk upfront

Answer + AI explanation with Pro

1. A.Multipathing (PSA/NMP)
2. B.qcow2
3. C.vSAN 8 ESA
4. D.RDM

Answer + AI explanation with Pro

1. A.RDM — mapping a physical LUN directly to a VM
2. B.RDM — Balances VMs across a datastore cluster by space usage and I/O latency, with affinity and anti-affinity rules.
3. C.RDM — A logical storage container in vSphere (VMFS, NFS or vSAN) that holds VM files, abstracting the underlying LUN or share
4. D.RDM — pre-allocating and zeroing a virtual disk upfront

Answer + AI explanation with Pro

1. A.VMware HCI software-defined storage pooling local disks
2. B.NVMe over Fabrics (RDMA, TCP, or FC) gives ESXi low-latency block storage with deeper queues than SCSI.
3. C.Allocating more virtual disk capacity than physically exists, relying on actual usage staying below the total; requires monitoring to avoid datastore-full outages
4. D.ESXi's pluggable storage architecture that manages multiple physical paths to a LUN for redundancy and load balancing (Round Robin, Fixed, MRU)

Answer + AI explanation with Pro

1. A.vSAN
2. B.iSCSI
3. C.Datastore
4. D.Raw Device Mapping (RDM)

Answer + AI explanation with Pro

1. A.vSAN — A mapping file on VMFS that gives a VM direct block access to a SAN LUN, used for clustering software or storage-array tools
2. B.vSAN — A logical storage container in vSphere (VMFS, NFS or vSAN) that holds VM files, abstracting the underlying LUN or share
3. C.vSAN — VMware HCI software-defined storage pooling local disks
4. D.vSAN — Tracks disk blocks modified since the last backup so incrementals copy only changes.

Answer + AI explanation with Pro

1. A.the architecture removing disk groups and a dedicated cache tier (all-NVMe)
2. B.vStorage APIs for Array Integration: offloads operations like full copy, block zeroing and hardware-assisted locking (ATS) to the storage array
3. C.ESXi's pluggable storage architecture that manages multiple physical paths to a LUN for redundancy and load balancing (Round Robin, Fixed, MRU)
4. D.Tracks disk blocks modified since the last backup so incrementals copy only changes.

Answer + AI explanation with Pro

1. A.RDM
2. B.vSAN 8 ESA
3. C.VAAI primitives
4. D.qcow2

Answer + AI explanation with Pro

1. A.vSAN 8 ESA — Telling the array which blocks a thin disk no longer uses so freed space returns to the pool; automatic on VMFS-6 at a configurable rate
2. B.vSAN 8 ESA — ESXi's pluggable storage architecture that manages multiple physical paths to a LUN for redundancy and load balancing (Round Robin, Fixed, MRU)
3. C.vSAN 8 ESA — the architecture removing disk groups and a dedicated cache tier (all-NVMe)
4. D.vSAN 8 ESA — vStorage APIs for Array Integration: offloads operations like full copy, block zeroing and hardware-assisted locking (ATS) to the storage array

Answer + AI explanation with Pro

1. A.an independent complete copy of a VM
2. B.the KVM copy-on-write disk format supporting snapshots
3. C.The number of outstanding I/Os an HBA or device can handle; too low throttles throughput, too high inflates latency.
4. D.An improved virtual disk managed independently of any VM, with its own identity and snapshots, used by CNS for Kubernetes persistent volumes

Answer + AI explanation with Pro

1. A.vSAN
2. B.qcow2
3. C.vVols and SPBM
4. D.Multipathing (PSA/NMP)

Answer + AI explanation with Pro

1. A.qcow2 — An improved virtual disk managed independently of any VM, with its own identity and snapshots, used by CNS for Kubernetes persistent volumes
2. B.qcow2 — The number of outstanding I/Os an HBA or device can handle; too low throttles throughput, too high inflates latency.
3. C.qcow2 — Allocating more virtual disk capacity than physically exists, relying on actual usage staying below the total; requires monitoring to avoid datastore-full outages
4. D.qcow2 — the KVM copy-on-write disk format supporting snapshots

Answer + AI explanation with Pro