Abstract: A first command is received from a workload, by a host system executing a nested container storage interface (CSI) driver, to create a persistent volume. The nested CSI driver converts the first command used by the workload to create the persistent volume to a second command used by the host system to increase a capacity of a previously generated persistent volume allocated to the workload. The second command is transmitted to an undercluster CSI driver of the host system, wherein the second command causes the undercluster CSI driver to increase the capacity of the previously generated persistent volume.

Abstract: A first command is received from a workload, by a host system executing a nested container storage interface (CSI) driver, to create a persistent volume. The nested CSI driver converts the first command used by the workload to create the persistent volume to a second command used by the host system to increase a capacity of a previously generated persistent volume allocated to the workload. The second command is transmitted to an undercluster CSI driver of the host system, wherein the second command causes the undercluster CSI driver to increase the capacity of the previously generated persistent volume.

Abstract: A first command is received from a workload, by a host system executing a nested container storage interface (CSI) driver, to create a persistent volume. The nested CSI driver converts the first command used by the workload to create the persistent volume to a second command used by the host system to increase a capacity of a previously generated persistent volume allocated to the workload. The second command is transmitted to an undercluster CSI driver of the host system, wherein the second command causes the undercluster CSI driver to increase the capacity of the previously generated persistent volume.

Abstract: Cloning of virtual-machine images can be managed. For example, a computing device can copy a segment of a virtual-machine image stored in a second storage device to a first storage device in response to receiving a first read request for the segment from a virtual machine. The first storage device may be capable of responding to read requests from the virtual machine with less latency than the second storage device. The computing device can also update a log to indicate that the segment is stored on the first storage device. Thereafter, the computing device can receive a second read request for the segment. In response, the computing device can determine that the segment is stored in the first storage device using the log, and provide the segment by obtaining the segment from the first storage device.

Abstract: Cloning of virtual-machine images can be managed. For example, a computing device can copy a segment of a virtual-machine image stored in a second storage device to a first storage device in response to receiving a first read request for the segment from a virtual machine. The first storage device may be capable of responding to read requests from the virtual machine with less latency than the second storage device. The computing device can also update a log to indicate that the segment is stored on the first storage device. Thereafter, the computing device can receive a second read request for the segment. In response, the computing device can determine that the segment is stored in the first storage device using the log, and provide the segment by obtaining the segment from the first storage device.