VCP-DCV (2V0-21.23) - Fact Sheet¶

Quick Reference¶

Exam Code: 2V0-21.23 Duration: 135 minutes Questions: 70 (multiple choice, single and multiple answer) Passing Score: 300 / 500 Cost: $250 USD

Official 2V0-21.23 Exam Prep Guide VMware Configuration Maximums Tool - authoritative source for all maximums VMware Compatibility Guide - HCL

vSphere 8 Configuration Maximums (high-yield)¶

Source: configmax.broadcom.com (vSphere 8.0 U3).

Per virtual machine¶

Per ESXi host¶

Per cluster¶

Note. vSphere 7 capped clusters at 64 hosts. The 96-host cap is a vSphere 8 increase. Some study material still says 64; trust configmax.broadcom.com.

Per vCenter Server¶

vSphere Distributed Switch (vDS)¶

VCSA Sizing (Tiny / Small / Medium / Large / X-Large)¶

Storage profiles add additional disk for /storage/seat (stats/events/alarms/tasks): Default, Large, X-Large.

License Editions and Subscription Bundles¶

Pre-Broadcom (legacy perpetual SKUs - still on questions you may see)¶

Post-Broadcom subscription bundles (Nov 2023 onward)¶

VMware retired most perpetual licenses and consolidated into two main subscription bundles:

VCF is licensed per-core (16-core minimum per CPU). VVF is also per-core. Standalone vSphere Standard is still per-CPU with 32-core minimum.

ESXi Boot Media Requirements (vSphere 8)¶

• Boot device must be at least 32 GB.
• USB / SD-card-only boot is deprecated: vSphere 8 still allows it but with persistent local disk for /scratch and tools, and shows warnings. New deployments should use a dedicated drive.
• ESX-OSData partition holds /scratch, system swap, ESX-OSData; sized 32-128 GB depending on disk size.
• M.2 SATA / NVMe SSD or local HDD/SSD recommended.

VM Hardware Versions (compatibility)¶

vmx-XX is the on-disk identifier. Older VMs run fine on newer hosts, but newer features need a HW upgrade.

esxcli Cheat Sheet¶

esxcli is the modern host-level CLI. Run on the ESXi shell or via SSH.

Host info¶

esxcli system version get
esxcli system hostname get
esxcli system maintenanceMode get
esxcli system maintenanceMode set --enable=true
esxcli hardware platform get
esxcli hardware cpu list
esxcli hardware memory get

Networking¶

esxcli network nic list                              # physical NICs (vmnicX)
esxcli network nic get -n vmnic0
esxcli network ip interface list                     # vmkernel ports (vmkX)
esxcli network ip interface ipv4 get
esxcli network vswitch standard list
esxcli network vswitch standard portgroup list
esxcli network vswitch standard add -v vSwitch1
esxcli network vswitch standard uplink add -u vmnic1 -v vSwitch1
esxcli network firewall ruleset list
esxcli network firewall ruleset set -e true -r sshServer

Storage¶

esxcli storage core adapter list
esxcli storage core device list
esxcli storage core path list
esxcli storage filesystem list                       # mounted datastores
esxcli storage vmfs extent list
esxcli storage nfs list
esxcli storage nfs add -H 10.0.0.5 -s /export/nfs01 -v NFS01
esxcli storage core device set -d <NAA.id> --state=on

vSAN¶

esxcli vsan cluster get
esxcli vsan cluster list
esxcli vsan storage list
esxcli vsan health cluster list
esxcli vsan policy getdefault

VM operations¶

vim-cmd vmsvc/getallvms                              # list VMs (use vim-cmd, not esxcli)
vim-cmd vmsvc/power.on <vmid>
vim-cmd vmsvc/power.off <vmid>
vim-cmd vmsvc/snapshot.create <vmid>
esxcli vm process list                                # running VM worlds
esxcli vm process kill --type=soft --world-id=<wid>

Software / patching¶

esxcli software vib list
esxcli software vib install -d /vmfs/volumes/datastore1/patch.zip
esxcli software profile update -p <profile> -d <depot>
esxcli software vib remove -n <vibname>

Logs¶

esxcli system syslog config get
esxcli system syslog config set --loghost='udp://10.0.0.50:514'
esxcli system syslog reload

PowerCLI Cheat Sheet¶

PowerShell module for vCenter / ESXi automation. Install: Install-Module VMware.PowerCLI.

Connect / disconnect¶

Connect-VIServer -Server vcenter.lab.local -User administrator@vsphere.local
Disconnect-VIServer -Server vcenter.lab.local -Confirm:$false
Get-VIServer

Inventory¶

Get-Datacenter
Get-Cluster
Get-VMHost
Get-VMHost esx01.lab.local | Format-List Name, Version, Build, ConnectionState, PowerState
Get-Datastore
Get-VirtualSwitch -VMHost esx01
Get-VDSwitch
Get-VDPortgroup

VM operations¶

Get-VM
Get-VM web01 | Select Name, NumCpu, MemoryGB, PowerState, Folder, ResourcePool
New-VM -Name web02 -Template "Ubuntu-22-Template" -Datastore ds01 -ResourcePool ProdPool
Start-VM -VM web01
Stop-VM -VM web01 -Confirm:$false
Restart-VMGuest -VM web01
Move-VM -VM web01 -Destination esx02.lab.local                   # vMotion
Move-VM -VM web01 -Datastore ds02                                  # Storage vMotion
New-Snapshot -VM web01 -Name "before-patch" -Memory -Quiesce
Get-Snapshot -VM web01
Remove-Snapshot -Snapshot $snap -Confirm:$false

Bulk reporting¶

Get-VM | Select Name, NumCpu, MemoryGB,
    @{N='UsedSpaceGB';E={[math]::Round($_.UsedSpaceGB,1)}},
    @{N='Host';E={$_.VMHost.Name}} |
    Export-Csv vm-inventory.csv -NoTypeInformation

Cluster / DRS / HA¶

Get-Cluster Prod | Select Name, HAEnabled, DrsEnabled, DrsAutomationLevel
Set-Cluster -Cluster Prod -DrsAutomationLevel FullyAutomated -Confirm:$false
New-DrsRule -Cluster Prod -Name "AffinityWeb" -KeepTogether -VM web01,web02

Host config¶

Get-VMHost esx01 | Set-VMHost -State Maintenance
Get-VMHost esx01 | Set-VMHost -State Connected
Get-VMHost | Get-VMHostNetworkAdapter -VMKernel

Networking - vSwitch (VSS) vs vSphere Distributed Switch (vDS)¶

vSwitch policy stack¶

• Security: Promiscuous mode, MAC address changes, Forged transmits (default Reject for the latter two on VSS, Accept on default vDS)
• Traffic shaping: Average bandwidth, peak, burst (egress only on VSS, ingress+egress on vDS)
• Teaming and failover: Load balancing (Route based on originating port ID, IP hash, source MAC, NIC load), active/standby/unused uplinks

NIC teaming load-balancing modes¶

Storage¶

VMFS versions¶

VMFS-6 adds: automatic UNMAP, 4K native (4Kn) sector support, SE Sparse improvements.

vSphere 7.0 U1+ dropped VMFS-3 support. vSphere 8.0 does not support VMFS-3. VMFS-5 is supported but VMFS-6 is the default for new datastores.

NFS versions¶

NFS v3 and NFS v4.1 datastores cannot be mounted by both protocols on the same host simultaneously.

vSAN concepts¶

• Hybrid vs all-flash. Hybrid uses SSD cache + HDD capacity. All-flash uses SSD cache + SSD capacity. All-flash supports dedupe and compression.
• Disk groups. 1 cache device + 1-7 capacity devices = 1 disk group. Up to 5 disk groups per host.
• OSA vs ESA. Original Storage Architecture (OSA) uses disk groups. Express Storage Architecture (ESA), new in vSphere 8, uses a flat NVMe-only "storage pool" with no cache tier. ESA needs vSAN ReadyNodes certified for ESA.
• Storage policies. FTT (Failures to Tolerate), FTM (RAID-1 mirror, RAID-⅚ erasure coding), stripe width, IOPS limit, force provision.
• Stretched cluster. Two sites + witness host for quorum.
• Hosts in vSAN cluster: 2-host (ROBO), 3-64 standard, 96 max in vSphere 8.

Datastore types comparison¶

High Availability (vSphere HA)¶

• Restarts VMs on surviving hosts after a host failure.
• Master/slave model. One host is the master, talks to vCenter and tracks the cluster.
• Heartbeats: management network + heartbeat datastores (default 2).
• Admission control: prevents over-commitment so the cluster can recover from N host failures. Default policy: Cluster resource percentage. Alternatives: Slot policy, Dedicated failover hosts.
• VM monitoring: restarts a VM if VMware Tools heartbeat stops.
• Application monitoring: requires SDK integration.
• Proactive HA: vendor health provider triggers maintenance migration before a host fails.

HA states for hosts¶

• Connected, Maintenance, Disconnected, Not Responding, Isolated, Network Partitioned

DRS (Distributed Resource Scheduler)¶

• Balances VM placement across hosts based on CPU/memory load.
• Automation modes: Manual, Partially Automated (placement only), Fully Automated (default).
• Migration threshold: 1 (conservative) to 5 (aggressive). Default 3.
• Predictive DRS (with Aria Operations) uses 24h forecasts.
• DRS rules: VM-VM affinity (keep together), anti-affinity (separate), VM-Host (must / should run on / not run on).

vMotion / Storage vMotion¶

• vMotion moves a running VM's CPU/memory state between hosts. Requires shared storage or vSphere 6.0+ with cross-vSwitch / cross-vCenter / cross-VC vMotion.
• Storage vMotion moves VM files between datastores while running.
• Long Distance vMotion: up to 150 ms RTT, requires 250 Mbps+ per concurrent vMotion.
• Encrypted vMotion: Required, Opportunistic (default), Disabled.

vMotion network requirements¶

• Dedicated VMkernel adapter with vMotion service enabled.
• 1 GbE supports 4 concurrent vMotions; 10 GbE supports 8.
• 25/40/100 GbE supported and recommended for high-density.

Fault Tolerance (FT)¶

• Lock-step secondary VM stays in sync with primary. Zero-downtime failover.
• vSphere 8 FT: up to 8 vCPUs and 128 GB RAM per FT VM.
• Max 4 FT-protected VMs per host, 8 vCPUs total per host.
• Requires dedicated 10 GbE FT logging network. Hardware Virtualization (HV) must be enabled.

Lifecycle Manager (vLCM)¶

Replaces vSphere Update Manager (VUM). Two modes:

• Baselines (legacy, vSphere 7-8): manage hosts with patch and upgrade baselines.
• Image-based (vSphere 7+, recommended): single desired-state image (ESXi base + vendor add-ons + components + firmware) applied cluster-wide.

vLCM also patches VMware Tools, VM hardware, and standalone hosts. Image-based clusters cannot mix vendors (e.g., all Dell, or all HPE).

Logs to know (Section 6)¶

Bundle support logs with vm-support on ESXi or "Export System Logs" in the vSphere Client.

esxtop counters (high-yield)¶

Press the listed key in esxtop to switch view.

Identity, RBAC, SSO¶

• vCenter SSO: domain default vsphere.local. First user administrator@vsphere.local.
• Identity sources: Integrated Windows Authentication, AD over LDAP, OpenLDAP, vsphere.local. AD over LDAPS recommended.
• Federated identity: vCenter 7.0+ supports OIDC federation with AD FS, Okta, PingFederate, Azure AD/Entra ID.
• Roles: Default - No access, Read-only, Administrator, plus sample roles (Virtual Machine Power User, Datastore Consumer, Network Administrator, etc).
• Permissions: assigned at object level (vCenter, datacenter, cluster, host, folder, VM, network, datastore). Permissions propagate down by default.
• Global permissions (across linked vCenters) vs vCenter-scoped permissions.

Tags and Categories¶

• Categories group related tags (e.g., "Environment" → Prod, Dev, QA).
• Cardinality: One tag per object, or many tags per object (per category).
• Used for: search, content library publishing, storage policies, DRS host groups, backup product policies.

Content Library¶

• Stores templates (OVF/OVA), ISOs, and other files. Replaces ad-hoc template folders.
• Local library vs Subscribed library (pull from a published one). Subscribed-on-demand vs Subscribed-immediate.
• Cross-vCenter publish/subscribe is supported.
• Storage backend: any datastore (VMFS, NFS, vSAN). NFS recommended for large libraries.

Common confusions¶

• VUM is gone. vSphere 8 uses vLCM. Don't pick "Update Manager" answers.
• PSC is gone. Platform Services Controller was retired with vSphere 7. SSO/CA/VMDir now run inside VCSA.
• External vCenter database is gone. VCSA uses an internal embedded vPostgres database. Windows vCenter Server was retired with vSphere 7.
• NFS v4.1 with Kerberos and VAAI: VAAI-NAS plugin needed for offloads.
• APD vs PDL. All Paths Down (APD) is transient; storage might come back. Permanent Device Loss (PDL) is final - array reports the LUN gone. HA reactions differ.
• HA vs FT. HA restarts VMs (downtime). FT runs a lockstep secondary (zero downtime, but performance cost).
• Memory ballooning vs swapping. Ballooning is benign reclaim via VMware Tools driver. Host swapping (.vswp) means real RAM exhaustion. Compression is in-between.
• Reservations vs Shares vs Limits. Reservation = guaranteed minimum. Limit = hard cap. Shares = priority during contention.

High-yield memorization list¶

• vSphere 8 maximums table above (per VM, per host, per cluster, per vCenter)
• VCSA sizing: Tiny / Small / Medium / Large / X-Large host counts
• License bundles post-Broadcom: VCF vs VVF vs vSphere Standard contents
• vSwitch teaming modes and when each is required
• VMFS-5 vs VMFS-6 differences (UNMAP, 4Kn)
• vSAN OSA disk groups (1 cache + 1-7 capacity, max 5 per host)
• vSAN ESA storage pool concept (vSphere 8 NVMe-only)
• HA admission control policies (Cluster resource percentage default)
• DRS automation modes and migration threshold default of 3
• FT vSphere 8 limits: 8 vCPU, 128 GB RAM, 4 VMs per host
• Logs by component (vmkernel, hostd, vpxa, vpxd, fdm)
• esxtop counters and thresholds (%RDY, DAVG, %SWP)