Home > Storage > PowerMax and VMAX > Storage Admin > Dell PowerMax and VMware vSphere Configuration Guide > Copying virtual machines with TimeFinder
Since TimeFinder/SnapVX provides the flexibility of creating copies of devices without specifying a target, it distinguishes it from all other TimeFinder technologies. This offers customers greater flexibility and functionality to make copies of production data.
The steps to create array copies are the same whether a particular virtual machine has a raw device mapping (RDM). However, RDM devices must be included with the VMFS devices when creating a copy to ensure the VM is complete. Failure to include the RDM devices will result in errors powering on the virtual machine after restoring the copy.
Note: TimeFinder/SnapVX technology also offers the option of copying the data when the target is linked by using the “COPY” option. Typically this operation should only be done if the linked target is going to be used for an extended period. In PowerMaxOS 10, all copies are NOCOPY.
Note: If the intended linked target device is in use as a VMFS datastore, the VMs should be unregistered. The datastore should then be unmounted from the ESXi hosts before undertaking the array snapshot process.
Devices are managed as an atomic unit by creating a device group, composite group, or using the storage group when taking a snapshot. Composite groups must be used if the virtual machines that are being cloned use VMFS or RDM volumes from multiple PowerMax storage arrays. Solutions Enabler or Unisphere for PowerMax can be used to create the device group and to manage the snapshot process. Only Solutions Enabler can create composite groups. Both methods are presented here for completeness.
Note: Virtual volume datastores or individual vVols cannot be copied with this process as there is no external snapshot management.
The environment for this example consists of:
The following presents the CLI steps that are required to copy a group of virtual machines using SnapVX technology. Before beginning the process, gracefully shut down any VMs running on the selected datastores as this shutdown is a cold process.
symdg create <group_name>
symdg -g <group_name> add dev <dev #> -sid <array SN>
Figure 129 demonstrates the commands and then outputs the content of the device group.
symdg -g <group_name> -sid <array SN> -tgt add dev <dev #>
symsnapvx -g <group_name> -name <snapshot name> establish
Unlike other TimeFinder technologies, you can take up to 256 snapshots of these groups of devices without linking them to a target (1024 for storage groups). In this example, however, the next step is to link the devices. Up to 1024 linked targets are supported.
symsnapvx -g <group_name> -snapshot_name <snapshot name> link
The following presents the UI steps that are required to copy a group of virtual machines using SnapVX technology. Before beginning the process, gracefully shut down any VMs running on the selected datastores.
Unisphere for PowerMax has other options available (which are also available through CLI). In this example in Figure 134, the Time To Live is set which indicates when the snapshot expires. Dell Technologies always recommends setting this option so resources are reclaimed when the snapshot is no longer needed.
Dell Virtual Storage Integrator (VSI) version 8.5 and above also offers some limited snapshot integration. Figure 136 shows one of the interfaces where, after taking a snapshot of a datastore, it can be linked to a new device that VSI will automatically create. For additional detail, see the VSI Product Guide.
VMware assigns a unique signature to all VMFS volumes when they are formatted with VMFS. Furthermore, the VMFS label is also stored on the device.
Since storage array technologies create exact replicas of the source volumes, all information including the unique signature and label is replicated. If a copy of a VMFS volume is presented to any VMware ESXi host or cluster group, the VMware ESXi, by default, automatically masks the copy. The device holding the copy is determined by comparing the signature that is stored on the device with the computed signature. The SnapVX targets, for example, have a different ID from the standard device it is associated with it. Therefore, the computed signature for a target device always differs from the signature that is stored on it. This differentiation enables VMware ESXi to always identify the copy correctly.
VMware vSphere individually resignatures or mounts VMFS volume copies by using either:
Since it is volume-specific, it allows for greater control in the handling of snapshots. This feature is useful when creating and managing volume copies that local replication products such as SnapVX, or remote replication technologies such as SRDF, create.
While either the CLI or the UI can be used to perform these operations, the UI (vSphere Client) is always recommended. Nevertheless, there are a few features available in the CLI that cannot be performed in vSphere Client.
In two of the above cases of creating SnapVX copies (CLI or UI), once the steps are complete the copied devices can be presented to a vCenter. When using VSI, it facilitates presentation to VMware automatically. In the following steps, a backup vCenter is used, however copies can be presented back to the original vCenter. In such cases, you must understand how duplicate extents work. An explanation can be found later in this chapter in Using TimeFinder copies.
The steps below use a linked target storage group that is named snapshot_VMFS_target_sg which has two 250 GB VMFS devices and one 10 GB RDM device.
Note: This scenario is true if the original device is not presented to the ESXi host, in which case the mount of the VMFS volume copy fails.
Dell Technologies recommends assigning a new signature to prevent any duplicate extents in the future. If the devices have been presented to a different vCenter, however, it is possible to keep the existing signature. It is unnecessary to type a new datastore name as VMware automatically generates one.
Note: After the new RDM is configured on the copied VM (using the same VMDK file name as the original VM), it may be useful to make a copy of the new vmx file. If the environment is refreshed, and only the VM data has changed and not the devices, the saved vmx file can be copied back to the refreshed environment. Doing so avoids having to re-add the RDM, and instead the VM can be immediately powered-on.
In the ESXi CLI, the esxcfg-volume (or vicfg-volume with the VMware remote CLI tools) command is the function that is used to resignature VMFS volume copies or mounts cloned ones. A feature that is unique to esxcfg-volume and not available in vSphere Client is the ability to nonpersistently force-mount a volume copy. The volume copy will not be mounted after a reboot of the ESXi host).
esxcfg-volume comes with the following parameter options:
-l | --list
-m | --mount <vmfs uuid|label>
-u | --umount <vmfs uuid|label>
-r | --resignature <vmfs uuid|label>
-M | --persistent-mount <vmfs uuid|label
The command esxcfg-volume -l lists all the volumes that have been detected as snapshots or replicas. If a volume copy is found, the command displays the following:
Figure 145 shows the list command returns two VMFS volume copies. The first volume copy cannot be mounted because the original volume is still online. The second volume can be mounted because the original VMFS volume is not present on that ESXi host.
In this example, the first discovered volume copy “62603e14-5f601c30-5f40-34800d0ed39c/STAR_341_ASYNC_302” is nonpersistently force-mounted using the command esxcfg-volume -m. Since it does not have its original VMFS volume present, it can be mounted without the requirement to resignature.
In Figure 146, this volume is nonpersistently mounted through the CLI. Then esxcfg-scsidevs -m is run to list the VMFS volumes. The newly mounted volume is listed. Since it is nonpersistent, it will disappear after the next reboot of the ESXi host.
Otherwise, if the cloned volume is never going to be presented to this host again it can be mounted persistently. This method is the equivalent of using the vSphere Client Add Storage wizard to do it, by using the esxcfg-volume -M command.