VMware Certified Technical Associate (VCTA) Study Guide
The VMware Certified Technical Associate (VCTA) - Data Center Virtualization is an entry-level credential validating foundational knowledge of VMware vSphere architecture and day-to-day administration tasks. It is aimed at IT operators, junior administrators, help-desk staff, and anyone beginning a vSphere career who needs to understand virtualization concepts and basic vSphere operations. The 135-minute exam (1V0-21.20) contains about 51 questions, is scored on a 100-500 scale with a passing score of 300, and assumes hands-on familiarity with vCenter and ESXi.
Domain 1: Virtualization Concepts
- VMware ESXi is a Type-1 (bare-metal) hypervisor that installs directly onto physical server hardware with no underlying host OS, abstracting CPU, memory, storage, and network for multiple VMs.
- A virtual machine is a software-defined computer with virtual CPU (vCPU), virtual memory, virtual disk (VMDK), and virtual NIC, isolated from the physical hardware and other VMs.
- vCenter Server is the centralized management platform that aggregates many ESXi hosts and unlocks enterprise features such as vMotion, HA, and DRS; ESXi and vCenter Server are the two foundational vSphere components.
- A vSphere cluster is a logical grouping of ESXi hosts whose CPU and memory are pooled to enable shared-resource features like High Availability (HA) and Distributed Resource Scheduler (DRS).
- Administration is done primarily through the vSphere Client, an HTML5 web interface that connects to vCenter Server; the legacy Flash-based Flex client and the C# desktop client are retired.
- Reservations guarantee a fixed minimum amount of CPU or memory to a VM at all times, while shares set relative priority only when resources are contended, and limits cap maximum usage.
- Transparent Page Sharing (TPS) deduplicates identical memory pages so multiple VMs reference a single physical copy, reducing overall memory consumption.
- Hypervisor (VMkernel) swapping to disk is the most severe memory-reclamation technique for performance and should be avoided during normal operation; ballooning and compression are preferred first.
- Heavy vCPU overcommitment increases CPU Ready time and CPU co-scheduling latency, so VMs should be sized with only the vCPUs the workload actually needs.
- vSphere Fault Tolerance (FT) provides zero-downtime failover via a continuously running secondary shadow VM, but it consumes a duplicate set of CPU, memory, and storage resources.
- A VMkernel adapter (vmk) carries host system traffic such as management, vMotion, vSAN, and provisioning; each service type can be assigned to its own VMkernel port.
- Host profiles capture a reference host's configuration so it can be applied to or audited against other hosts to enforce configuration consistency across a cluster.
- Virtualization drives higher hardware utilization through consolidation, which reduces the number of physical servers and lowers power and cooling costs while speeding VM provisioning.
- Advanced features (vMotion, HA, DRS, FT) require higher vSphere editions, so edition choice directly drives licensing cost.
- Useful esxcli hardware commands: 'esxcli hardware cpu list' shows per-CPU core/thread topology, 'esxcli hardware memory get' reports total/available RAM, and 'esxcli hardware platform get' plus 'esxcli system version get' return host hardware and build details.
Domain 2: vSphere Management
- A VM snapshot is a point-in-time capture of a VM's disk, configuration, and optionally memory that you can revert to; it is for short-term use, not as a backup.
- Snapshots are stored as delta disk files on the same datastore as the VM, grow continuously as the VM writes, and are lost if the datastore fails, so they do not protect against datastore loss.
- VMware Tools is a suite of paravirtualized drivers and utilities installed in the guest OS that improves performance, enables graceful shutdown and quiesced snapshots, and supports host-guest integration.
- Paravirtualized devices VMXNET3 (network) and PVSCSI (storage) require VMware Tools and deliver the best performance with the lowest CPU overhead.
- A VM's core configuration lives in the .vmx file, with virtual disks in .vmdk files; together they define vCPU, memory, NIC, and disk settings.
- A template is a read-only master image of a VM used to deploy consistent new VMs quickly; a clone is an exact copy of an existing VM (full clones are independent, linked clones share base disk).
- Guest customization (a customization specification) assigns unique identity such as hostname, IP address, and SID when deploying VMs from a template, avoiding identity conflicts.
- Deploying an OVF or OVA template provisions a preconfigured VM (or appliance) from a packaged, portable file.
- vCenter Server is delivered as the vCenter Server Appliance (VCSA), a preconfigured Linux-based virtual appliance running on Photon OS; the Windows-installable vCenter is discontinued.
- Role-Based Access Control (RBAC) assigns privileges through roles attached to users or groups on inventory objects, applying the principle of least privilege in vCenter.
- Integrating vCenter with an identity source such as Active Directory through vCenter Single Sign-On (SSO) centralizes authentication and protects access.
- Running guest TRIM/UNMAP reclaims deleted blocks on a thin-provisioned datastore, returning unused capacity to the storage.
- Stale or forgotten snapshots and orphaned VMs consume datastore capacity and backup scope without delivering value and should be cleaned up.
- The esxcli command 'esxcli vm process list' lists running VMs with their World IDs, which can then be used with 'esxcli vm process kill' to forcibly stop an unresponsive VM.
Domain 3: Storage and Networking
- A virtual switch (vSphere Standard Switch or Distributed Switch) is a software Layer 2 switch in ESXi that connects VM and VMkernel ports to each other and to the physical network through uplinks (vmnic).
- A port group is a named policy object on a virtual switch that defines VLAN ID, security settings (promiscuous mode, MAC changes, forged transmits), and traffic shaping for the ports within it.
- VLAN tagging on a port group isolates VM traffic at Layer 2 by assigning a VLAN ID to all VMs connected to that group.
- A datastore is the logical storage container where VM files (.vmx, .vmdk, snapshots) reside, abstracting underlying physical storage.
- VMFS is VMware's clustered filesystem on block storage (Fibre Channel or iSCSI LUNs) allowing concurrent multi-host access; NFS datastores mount file shares from a NAS over the network.
- VMware vSAN is software-defined, hyper-converged storage that pools the local flash and disk devices of cluster hosts into a single shared datastore governed by storage policies.
- vSAN and vMotion traffic typically each require a dedicated VMkernel network adapter, and vMotion needs both shared storage access and a vMotion network to move VMs between hosts.
- Thin provisioning allocates datastore space on demand as the guest writes data, whereas thick provisioning (lazy or eager zeroed) reserves the full disk capacity up front.
- NIC teaming combines multiple physical uplinks on a virtual switch to provide redundancy and load distribution if a NIC or path fails.
- VMXNET3 is the recommended paravirtual virtual NIC, providing the best network throughput with low CPU overhead.
- Storage DRS automatically migrates VM files between datastores in a datastore cluster to balance space utilization and I/O latency.
- Storage I/O Control (SIOC) provides cluster-wide I/O prioritization using disk shares when a datastore experiences congestion, and Network I/O Control (NIOC) allocates bandwidth among traffic types using shares and limits.
- Storage tiering practice: place latency-sensitive VMs on fast flash/SSD-backed datastores and archival or low-I/O VMs on cheaper, higher-capacity storage.
- Monitor capacity with datastore usage alarms and compare actual consumption against provisioned capacity, especially when overcommitting with thin disks.
Domain 4: VM Operations
- vMotion performs live migration of a running VM's active memory and execution state between ESXi hosts over a dedicated vMotion network with no downtime.
- Storage vMotion migrates a running VM's files between datastores live, without interrupting the VM, enabling storage maintenance and rebalancing.
- vSphere HA elects a master host that monitors slaves via management-network and datastore heartbeats, and automatically restarts a failed host's VMs on surviving hosts.
- DRS continuously evaluates cluster CPU and memory load and, in fully automated mode, uses vMotion to migrate VMs to balance the cluster without administrator approval.
- Resource pools are logical containers that hierarchically partition cluster CPU and memory among groups of VMs using shares, reservations, and limits.
- vSphere Fault Tolerance runs a lockstep secondary shadow VM so that if the primary host fails the secondary takes over instantly with no restart and no data loss.
- Shared storage is the prerequisite that makes a VM's files reachable by all hosts, enabling vMotion, HA, and DRS.
- Distributed Power Management (DPM) consolidates VMs onto fewer hosts during low demand and powers down idle hosts to save energy, powering them back on when load rises.
- Proactive HA evacuates VMs from a host that a hardware health provider predicts will fail, moving them before an actual failure occurs.
- High CPU Ready time and high CPU Co-stop are the key counters indicating a CPU scheduling problem, often caused by too many vCPUs; reduce vCPUs to match the parallel workload.
- A high shares value plus a CPU and memory reservation guarantees performance for a business-critical VM under contention.
- Each active snapshot creates a delta disk that grows and adds read/write overhead as the chain lengthens, so long snapshot chains degrade performance and should be consolidated.
- HA admission control reserves failover capacity; the percentage-based policy should be sized to actual failover need and host failures to tolerate set to match real redundancy requirements.
- Editing a VM's virtual hardware settings adjusts CPU and memory allocation, and converting a VM to a template lets you deploy consistent clones rapidly from a master image.
VMware Certified Technical Associate (VCTA) exam tips
- Memorize which features require vCenter versus a standalone ESXi host: vMotion, HA, DRS, FT, and DPM are all cluster features that depend on vCenter Server, while a lone ESXi host can only run and manage its own local VMs.
- Know the memory-reclamation order and severity cold: TPS and ballooning are lightweight, memory compression is moderate, and hypervisor swapping to disk is the last resort and worst for performance.
- Distinguish reservation (guaranteed minimum), limit (hard maximum), and shares (relative priority only during contention) because resource-allocation questions hinge on these exact definitions.
- Do not confuse snapshots with backups: snapshots live on the same datastore, grow over time, and are lost with the datastore, so they are never a disaster-recovery solution.
- Match each storage type to its transport: VMFS sits on block storage (FC/iSCSI), NFS is file storage over the network, and vSAN pools local host disks; expect questions that test these mappings.
Study guide FAQ
What score do I need to pass the VCTA exam and how long is it?
The exam is scaled and requires a passing score of 300 on a 100-500 scale. You get 135 minutes to answer roughly 51 multiple-choice questions, and there is no negative marking, so answer every question.
Do I need hands-on vSphere experience or is studying theory enough?
The VCTA is an associate-level credential focused on concepts and basic operational tasks, so deep hands-on experience is not strictly required. However, even brief lab time in the vSphere Client (creating a VM, taking a snapshot, viewing datastores and networks) makes the terminology far easier to retain.
What is the difference between vSphere HA and Fault Tolerance, since both protect availability?
vSphere HA restarts a failed host's VMs on surviving hosts, which means a short outage while the VMs reboot. Fault Tolerance runs a lockstep secondary shadow VM that takes over instantly with zero downtime and zero data loss, but it consumes a full duplicate set of CPU, memory, and storage resources.
Is ESXi the same thing as vCenter Server?
No. ESXi is the Type-1 hypervisor installed on each physical server that actually runs the virtual machines. vCenter Server is a separate management appliance (the VCSA) that connects to many ESXi hosts to provide centralized administration and to enable cluster features like vMotion, HA, and DRS.