Home > Storage > PowerStore > Virtualization and Cloud > Dell PowerStore: Microsoft Hyper-V Best Practices > Application consistency with snapshots
PowerStore snapshots taken without regard to the state of the host operating system or workload are crash consistent. PowerStore snapshots taken manually, or as part of a recurring schedule, are crash-consistent by default. Administrators must place a host or workload in a consistent (paused) state if a consistent snapshot is essential for recovery.
If you use a crash-consistent snapshot to restore a host, VM, or workload, it is like starting the resource after an unexpected event such as a power outage.
Often, host operating systems and nontransactional workloads will recover to a crash-consistent state without complications. However, there is an elevated risk of data loss or corruption if you attempt to recover a transactional workload such as Microsoft SQL Server to a crash-consistent state.
Consider these recommendations if you need host or workload consistency before taking a PowerStore snapshot.
After the host or workload is in a consistent state (paused), take a PowerStore snapshot. Then, return the host or workload to its active state.
If possible, use scripting and automation tools to orchestrate a workflow that performs these steps automatically.
The following sections provide examples for how to use PowerStore snapshots and thin clones in a Hyper-V environment.