Home > Storage > PowerStore > Virtualization and Cloud > Dell PowerStore: VMware Site Recovery Manager Best Practices > Snapshots and application consistency
Asynchronous replication uses snapshots to provide point-in-time images as the source of RPO-based updates to the destination. These snapshots are used to maintain the common base images between the source and replicated resource across systems. Snapshots that replication creates and maintains are not visible in PowerStore Manager. When replication is configured, any snapshots created on the source resource are automatically replicated in chronological order to the destination system during the next RPO-based update (see Figure 8). There are several methods available for creating snapshots: PowerStore Manager, protection policies, PowerStore REST API, and PSTCLI. When replicated, SRM may use a thin clone of the snapshot to present recovered data to the vSphere cluster. Snapshots created in PowerStore Manager, or a protection policy are considered crash consistent. You can use other methods that result in application consistency within the snapshot. For example, where supported, you can use Dell Technologies AppSync to create application-consistent snapshots. This practice ensures that all incoming I/O for a given application is quiesced and flushed before a snapshot is created. Another method is to use vSphere snapshots with quiescence captured inside a replicated PowerStore snapshot. Either of these examples results in application-consistent snapshots being replicated to the recovery site.
When using vSphere snapshots, there are two important facts to recognize:
When the SRM recovery plan workflow is carried out, SRM registers the VM into inventory at the destination site. Then, it powers on the VM with no special attention given to the current snapshot state of the VM. This means that SRM powers on the VM using the delta, resulting in recovery from a crash-consistent state. To recover the VM from the frozen-parent disk with application and data consistency, revert the VM to the previous snapshot using the vSphere Snapshot Manager before powering on the VM. Once this process is done, you can delete (close) the snapshot and power on the VM. This process ensures the VM is powered on from its frozen-parent disk and the delta disk, and the crash-consistent data in it is destroyed.
If manually carrying out the previous process on a large scale, this can erode efforts made toward meeting the recovery plan RTO and is not the best use of SRM. In such instances, it is more efficient and consistent to script the snapshot management process using Microsoft PowerShell. You can carry out this process as a pre-power-on step (or potentially post-power-on step) for the VM using a custom recovery task.