VMware High Availability and Snapshots
VMware virtualization offers powerful features that keep applications running smoothly and protect data. Two of the most critical capabilities are High Availability (HA) and snapshots . This…

In a Fault Tolerance (FT) configuration, which limitation applies to the virtual machine's CPU resources?
When a snapshot is active, how is new data written to the original disk handled?
Which of the following best describes the impact of a storage‑only snapshot that does not include RAM and CPU state when rolled back?
What prerequisite is shared between High Availability (HA) and Fault Tolerance (FT) configurations?
During a fault‑tolerant failure, if the primary VM fails, what is the immediate action taken by the FT mechanism?
Which scenario best illustrates a use case for VM snapshots rather than traditional backups?
What is a key limitation of software‑only snapshots compared with hardware‑assisted snapshots?
In the context of vSphere HA, what happens to VMs on a host that becomes isolated from the cluster?
Why is it recommended to treat snapshots as short‑term tools rather than long‑term backups?
Which component ensures data consistency for a running VM during a backup operation?
What is the effect of enabling vSphere Fault Tolerance on CPU and RAM resources for the protected VM?
During a disaster recovery replication setup, why are snapshots preferred over full VM copies for replication?
Which of the following statements accurately reflects a challenge when configuring Fault Tolerance for a VM?
What is the primary benefit of vSphere Distributed Resource Scheduler (DRS) in a cluster?
In a vSphere HA cluster, what determines the order in which VMs are restarted after a host failure?
Understanding VMware High Availability (HA) and Snapshots
VMware virtualization offers powerful features that keep applications running smoothly and protect data. Two of the most critical capabilities are High Availability (HA) and snapshots. This course explains the core concepts, how they work together, and the practical limits you need to know when designing resilient virtual environments.
What Is a High Availability (HA) Cluster?
In a virtualized data center, a HA cluster is a group of ESXi hosts that share common storage and cooperate to protect virtual machines (VMs) from host‑level failures. The primary function of HA is to automatically restart virtual machines on another host after a failure. When a host crashes, the cluster’s master node detects the outage, selects a healthy host, and powers on the affected VMs based on predefined restart priorities.
- HA monitors host heartbeats and VM health checks.
- It uses resource pools to balance CPU and memory when restarting VMs.
- HA does not synchronize VMs for continuous operation; it only provides rapid recovery.
Fault Tolerance (FT) – Continuous Availability
While HA recovers after a failure, Fault Tolerance (FT) offers true continuous availability. FT creates a secondary VM that runs in lockstep with the primary. If the primary VM fails, the secondary instantly takes over without any downtime.
However, FT has a notable limitation on CPU resources: the virtual machine can have only one or two virtual CPUs. This restriction ensures that the lockstep synchronization can be maintained across the two VMs without overwhelming the host’s networking and processing capabilities.
Snapshots: Capturing a Point‑in‑Time State
A snapshot records the state of a VM at a specific moment. When a snapshot is active, the original virtual disk is marked read‑only, and all subsequent writes are redirected to a delta (or child) disk. This mechanism allows you to revert to the snapshot later while preserving the original data.
- The snapshot consists of three components: disk state, memory state, and CPU register state.
- When only the disk state is captured (a storage‑only snapshot), rolling back will power the VM on in a powered‑off state. The VM must be manually started, and any in‑memory data is lost.
- Including memory and CPU state creates a full snapshot, which can resume the VM exactly where it left off.
Shared Storage: The Common Prerequisite for HA and FT
Both HA and FT rely on shared storage accessible by all ESXi hosts. Shared storage ensures that:
- VM files (including snapshots) are visible to any host that may need to restart or take over a VM.
- FT can keep the primary and secondary VMs synchronized because they read and write to the same datastore.
- HA can quickly locate the VM’s configuration files during a host failure.
Note that the number of hosts required for HA or FT differs: HA can operate with as few as two hosts, while FT typically needs at least two hosts to host the primary and secondary VMs.
Fault‑Tolerant Failure Recovery Process
When a fault‑tolerant VM experiences a primary failure, the FT mechanism instantly promotes the secondary VM to become the new primary. This promotion happens without any pause or manual intervention, allowing the workload to continue uninterrupted. The former primary VM is automatically recreated on another host to maintain the FT pair.
When to Use Snapshots vs. Traditional Backups
Snapshots are ideal for short‑term, rapid‑recovery scenarios such as testing a new application version before deployment. They allow administrators to revert to a known good state within minutes, making them perfect for development, QA, and troubleshooting.
In contrast, traditional backups are designed for long‑term retention, disaster recovery, and compliance. Backups copy data to a separate location and can be stored for months or years, whereas snapshots are stored on the same datastore and are not a substitute for a proper backup strategy.
Software‑Only Snapshots vs. Hardware‑Assisted Snapshots
Software‑only snapshots are created by the hypervisor and involve writing changes to a delta file on the host. This approach introduces additional read/write overhead because every I/O operation must be duplicated to the delta disk. While flexible, this overhead can impact performance for I/O‑intensive workloads.
Hardware‑assisted snapshots, often provided by storage arrays, offload the snapshot process to the storage hardware, reducing host CPU usage and I/O latency. However, they require compatible storage solutions and may involve additional licensing costs.
Key Takeaways
- HA automatically restarts VMs on another host after a failure, providing rapid recovery.
- FT offers continuous availability but limits VMs to one or two virtual CPUs.
- When a snapshot is active, the original disk becomes read‑only and writes go to a delta disk.
- Storage‑only snapshots revert the VM to a powered‑off state; full snapshots preserve memory and CPU state.
- Both HA and FT require shared storage accessible by all hosts.
- During an FT failure, the secondary VM instantly becomes the primary.
- Snapshots are best for short‑term testing and rollback, not for long‑term backup.
- Software‑only snapshots add read/write overhead compared to hardware‑assisted solutions.
Frequently Asked Questions (FAQ)
Can HA and FT be used together on the same VM?
Yes, a VM can be protected by both HA and FT. HA provides host‑level recovery, while FT ensures zero‑downtime continuity. However, FT’s CPU limitations may restrict its use on larger VMs.
How many snapshots should I keep on a production VM?
Best practice recommends limiting active snapshots to one or two per VM. Keeping many snapshots can degrade performance and increase storage consumption.
Do snapshots replace backups?
No. Snapshots are a short‑term protection mechanism. For data durability, compliance, and disaster recovery, regular backups to a separate location are essential.
