If you run out of disk space, you can seamlessly increase the virtual disks of virtual machines that are configured for replication, without triggering an initial full synchronization.

You can resize the virtual disk of a replicated virtual machine while the virtual machine is powered on or powered off.
Important: As a best practice, resize the virtual disk of a replicated virtual machine while the virtual machine is powered on. If you increase the disk size of a virtual machine that is powered off, full synchronization is performed the next time you power on the replicated virtual machine.

After you increase the virtual disk on the source site, the virtual disk on the target site automatically resizes and the ongoing replication enters Resizing disk state, until the task completes.

Note: After you resize the virtual disk, vSphere Replication clears all available multiple points in time. You can modify this behavior by changing the virtual disk resizing configuration options.

To use this feature, you need vSphere 7.0 or later on the source site and vSphere 6.5 or later on the target site.

For more information about disk resizing, see Change the Virtual Disk Configuration in the VMware Host Client in the vSphere Product Documentation.

Configure the Virtual Disk Resizing

You can determine the behavior of vSphere Replication during disk resizing, by choosing one of the three configuration options. To activate your preferred option, you must change the value of three different parameters in the /etc/vmware/hbrsrv.xml configuration file.

  • vSphere Replication can follow two approaches to perform the disk resizing on the target site. To configure the way the server handles the resizing, change the value of the extendDiskPITHierarchyPolicy parameter.
    Table 1. extendDiskPITHierarchyPolicy Parameter Values
    Value Description
    extendDiskPITHierarchyPolicy = auto vSphere Replication selects preserve or collapse, depending on the current datastore storage consumption and the requested new virtual disk size. This is the default value of the parameter.
    extendDiskPITHierarchyPolicy = collapse vSphere Replication collapses the disk hierarchy of the replica disk and extends the resulting base disk. All PITs created before the start of the virtual disk resizing are lost. You cannot perform recovery until you create a PIT after resizing the virtual disk.
    extendDiskPITHierarchyPolicy = preserve vSphere Replication creates a new base disk, which is a full clone of the latest PIT. vSphere Replication extends the new disk to the new size. The original base disk still exists. The extra consumed storage is freed, after vSphere Replication removes all PITs, which contain the original disk. Then the vSphere Replication removes the original replica base disk.
  • To adjust the behavior when the extendDiskPITHierarchyPolicy is set to auto, you can use the extendDiskPITHierarchyPolicyAutoThreshold parameter. You can change the property value to а number between 0 and 1 (the default value is 0.9). This way you set a limit to the datastore capacity. vSphere Replication calculates this limit by multiplying the size of the datastore capacity by the extendDiskPITHierarchyPolicyAutoThreshold parameter value.

    For example, if the datastore capacity is 5 TB and the extendDiskPITHierarchyPolicyAutoThreshold parameter is set to 0.8, then the datastore capacity limit is 4 TB.

    vSphere Replication calculates what is the final storage consumption, if the preserve mode is active. If the storage consumption is below the threshold, vSphere Replication uses the preserve mode and if it is above the threshold, it uses the collapse mode.

  • To reduce the period of extended storage consumption, when extendDiskPITHierarchyPolicy is set to preserve mode, change the value of the removeMPITsBeforeBaseDisks parameter.
    Table 2. removeMPITsBeforeBaseDisks Parameter Values
    Value Description
    removeMPITsBeforeBaseDisks = true The vSphere Replication server drops all PITs, which are based on the original disk size, after a new PIT which is based on the extended disk appears.
    removeMPITsBeforeBaseDisks = false The retention policy of the PITs determines the expiration of the older PITs. The storage consumption drops, after all PITs, which refer to the original disk, are expired.