VMware Snapshot Best Practices Guide (2025)

A VMware snapshot is a short-term rollback tool, not a long-term protection or backup mechanism. The short answer is this: use snapshots only for short rollback windows before controlled changes, remove them as soon as validation is complete, leave enough free space for consolidation, and avoid using them carelessly on high-I/O or unsupported disk configurations. This guide is written for teams operating in the September 22, 2025 context.

Quick Summary

Broadcom KB 426571 says snapshots, especially for database workloads, are not backups and should be kept only for minutes to hours.
The same KB says long-running snapshots can impact VM and datastore performance and increase consolidation risk.
Broadcom KB 316414 recommends leaving at least 1.5x the total snapshot data size as free datastore capacity before consolidation, and says to free space first if usage is already at 99% or more.
The same KB says snapshot consolidation should never be interrupted, because it can risk irreversible virtual disk corruption.
Broadcom KB 338534 explains that consolidation time depends mainly on the aggregated delta size, snapshot chain depth, and snapshot overhead.
Broadcom KB 311074 says snapshots are not supported on physical mode RDM or on disks using bus-sharing.
Broadcom KB 412396 says backup and vSphere Replication can conflict because two snapshot-based processes cannot run on the same VM at the same time; the RPO must be longer than the backup duration or replication should be paused.
In the September 22, 2025 context, Broadcom KB 326316 lists vCenter Server 8.0 Update 3g / 8.0.3.00600 / Build 24853646 as one visible current vCenter 8 baseline.

What Is the Right Mindset for Snapshots?
Which vCenter Baseline Makes Sense on September 22, 2025?
When Should You Use a Snapshot?
What Are VMware Snapshot Best Practices?
How Should You Plan Consolidation Time and Space?
When Should You Avoid Snapshots?
Most Common Snapshot Operational Mistakes
First 15-Minute Checklist
Frequently Asked Questions

Server room image for a VMware snapshot best practices guide

Image: Wikimedia Commons - Kolos, le plus grand data center du monde .

What Is the Right Mindset for Snapshots?

A snapshot is a short rollback point. It is useful before patching, agent installation, configuration changes, or other controlled maintenance steps. It should not be treated like backup, archive, or long-term protection.

Broadcom KB 426571 is explicit about this for database virtual machines:

snapshots are not database-aware backups
delta disks can grow quickly on high-I/O workloads
long-lived snapshots increase performance and consolidation risk

The policy mindset should therefore be: create, validate, remove.

Which vCenter Baseline Makes Sense on September 22, 2025?

Snapshot operations are mainly driven by ESXi and storage behavior, but the operating standard should still be considered within the vCenter baseline. According to Broadcom KB 326316, one visible vCenter 8 line in the September 22, 2025 context is:

Product: vCenter Server 8.0 Update 3g
Version: 8.0.3.00600
Release date: 2025-07-29
Build: 24853646

This guide uses vCenter Server 8.0 Update 3g / Build 24853646 as the management baseline.

When Should You Use a Snapshot?

Good snapshot use cases are short and specific:

before operating system or middleware patching
before application configuration changes
before short rollback-window maintenance
before controlled test and validation steps

Broadcom KB 426571 says snapshots should be kept for minutes to hours only. Once the change is validated, the snapshot should be removed instead of left behind.

Related guides:

What Are VMware Snapshot Best Practices?

1. Keep Snapshots Short-Lived

This is the primary rule. A snapshot should not remain in place for days or weeks. The longer it stays open, the larger the delta growth and the larger the consolidation risk.

2. Do Not Confuse Snapshot with Backup

Broadcom KB 426571 explicitly says snapshots are not backups. This matters even more on SQL, Oracle, or other high-write database workloads.

3. Plan Free Space for Consolidation

Broadcom KB 316414 recommends leaving at least 1.5x the size of the total snapshot data as free datastore space before consolidation. The same source says that if the datastore is already at 99% usage or higher, free up space first.

That means deleting a snapshot is not simply a free-space recovery event. In many cases, consolidation needs working space before it can finish.

4. Never Interrupt Consolidation

Broadcom KB 316414 explicitly says snapshot consolidation should not be interrupted. Canceling or interfering at the wrong moment can lead to virtual disk corruption.

5. Avoid Building Deep Chains During Heavy I/O

Broadcom KB 338534 says consolidation time is driven by:

the aggregated size of the delta files
the snapshot chain depth
the size of snapshot overhead

Operationally, this means “we will remove it later” becomes riskier as the chain gets deeper and the delta files get larger.

6. Do Not Overlap Backup and Replication Snapshot Windows

Broadcom KB 412396 explains that vSphere Replication and third-party backup tools can conflict because two snapshot-based operations cannot run at the same time on the same VM. The fix is to make the RPO longer than the backup duration or pause replication during backup.

How Should You Plan Consolidation Time and Space?

Consolidation time should not be guessed from snapshot count alone. Broadcom KB 338534 ties it mainly to delta size and chain depth. Broadcom KB 316414 also points to an approximate 2 MB/s to 4 MB/s I/O rate range for estimation purposes.

The operational rule is straightforward:

do not plan the maintenance window before measuring snapshot size
do not start deletion before checking free datastore capacity
on critical VMs, estimate consolidation in the context of actual workload intensity

Especially on write-heavy systems, “the snapshot probably is not that large” is a bad assumption.

When Should You Avoid Snapshots?

Broadcom KB 311074 clearly lists configurations where snapshots are not supported:

Physical mode Raw Device Mappings
disks using bus-sharing

That means some cluster or special storage designs cannot rely on snapshots for rollback at all. In those cases, rollback and protection strategies need a different design.

Database workloads also need extra caution. Even when snapshots are technically possible, they should still be treated as short operational tools rather than a protection strategy.

Most Common Snapshot Operational Mistakes

Leaving snapshots open for days

This is the most common operational mistake. Broadcom KB 426571 says snapshots should remain short-lived. Long-lived snapshots unnecessarily increase performance and consolidation risk.

Trying to delete snapshots on a nearly full datastore

Broadcom KB 316414 shows that consolidation needs free space. The assumption that “deleting the snapshot will free space immediately” is not always valid.

Running backup and replication in the same window

Broadcom KB 412396 says that if two snapshot-based workflows collide, the backup may fail to start or fail to complete. The timing windows must be separated.

Expecting snapshots on unsupported disk topology

Broadcom KB 311074 is explicit: physical mode RDM and bus-sharing disk configurations do not support snapshots.

First 15-Minute Checklist

The purpose of the snapshot was confirmed as rollback, not protection
The removal time was decided before creation
Free datastore capacity was reviewed for consolidation
High-I/O or database workload impact was evaluated
Backup and replication windows do not overlap
No unsupported RDM or bus-sharing disk design exists
The snapshot will be removed as soon as validation finishes
The maintenance window includes realistic consolidation time

Next Step with LeonX

Snapshot policy is not just a technical setting. It needs to be tied to change management, backup windows, datastore capacity planning, and application ownership. LeonX helps teams turn snapshot usage into a controlled production standard.

Related pages:

Frequently Asked Questions

Can a snapshot replace backup?

No. Broadcom KB 426571 explicitly says snapshots are not backups. Database workloads in particular require application-aware backup methods.

How long should a snapshot be kept?

Broadcom KB 426571 says snapshots should be kept only for minutes to hours and removed as soon as the operation has been validated.

Why can deleting a snapshot require extra free space?

Because Broadcom KB 316414 says consolidation may require at least 1.5x the total snapshot data size as available free space.

Why does consolidation sometimes take a long time?

Broadcom KB 338534 says it depends on delta size, snapshot chain depth, and overhead size.

Which disk types do not support snapshots?

Broadcom KB 311074 says physical mode RDM and bus-sharing disk configurations do not support snapshots.

Conclusion

The core of VMware snapshot best practices is simple: use snapshots as short-term rollback tools, not as backup; leave room for consolidation; avoid unsupported disk designs; and do not let backup and replication compete for the same VM at the same time. In the September 22, 2025 context, the safest approach is a short-lived snapshot policy, realistic consolidation planning, and clear operational ownership.

Sources

Share this article

Facebook

Twitter