Abstract: A technique utilizes grafting and differential based (diff-based) data seeding to hydrate a special virtual disk (vdisk) on a multi-node cluster with data changes (differences) between a reference vdisk stored on the cluster and a snapshot stored in an external repository to enable failover (including failback) recovery of an application workload in a disaster recovery environment. The application workload is stored as a workload vdisk on local storage of the cluster and snapshots of the workload vdisk are generated and organized as a vdisk chain on the cluster. One or more snapshots of the vdisk chain may be replicated to the external repository using a long-term snapshot service. Each replicated snapshot may be backed up from the cluster to the external repository at the granularity of a vdisk, referred to herein as an external datasource disk. The special vdisk is a thinly provisioned, datasource-backed vdisk that is grafted onto the vdisk chain, e.g., as a child vdisk of the reference vdisk.

Abstract: A bypassing technique bypasses an indexing service and provides a bypass data path for transferring/retrieving snapshots from a production cluster to an object store. In an embodiment, the production cluster may determine how extents of the snapshots are packed into objects of the object store and transfers the snapshots directly to the object store over the bypass data path. Once the snapshot transfer is completed, the production cluster provides location metadata as to how the snapshot extents are packed into objects to the indexing service. The indexing service is invoked to create an index of the location metadata and is not involved in the data transfer of the snapshots. In another embodiment, the production cluster identifies a snapshot to restore and queries the indexing service to compute the deltas between the snapshot to be restored and a reference snapshot. The indexing service returns a set of segments that indicates the changed delta regions between the two snapshots.

Abstract: A technique utilizes grafting and differential based (diff-based) data seeding to hydrate a special virtual disk (vdisk) on a multi-node cluster with data changes (differences) between a reference vdisk stored on the cluster and a snapshot stored in an external repository to enable failover (including failback) recovery of an application workload in a disaster recovery environment. The application workload is stored as a workload vdisk on local storage of the cluster and snapshots of the workload vdisk are generated and organized as a vdisk chain on the cluster. One or more snapshots of the vdisk chain may be replicated to the external repository using a long-term snapshot service. Each replicated snapshot may be backed up from the cluster to the external repository at the granularity of a vdisk, referred to herein as an external datasource disk. The special vdisk is a thinly provisioned, datasource-backed vdisk that is grafted onto the vdisk chain, e.g., as a child vdisk of the reference vdisk.

Abstract: Various embodiments set forth techniques for transactional allocation and deallocation of blocks in a block store. A first technique includes sending a first request that causes a non-persistent allocation of a block. The first technique also includes adding a first entry in a log recording the allocation as tentative, sending a second request that causes persistence of the allocation, and adding a second entry in a log recording the allocation as finalized. A second technique includes adding a first entry in a log recording a deallocation of a block, sending a first request that causes the deallocation of the block and causes the block to be unavailable for reallocation in a non-persistent manner, adding a second entry in the log recording that the deallocation is finalized, and sending a second request that causes the block to be made available for reallocation.

Abstract: Various embodiments set forth techniques for free space management in a block store. The techniques include receiving a request to allocate one or more blocks in a block store, accessing a sparse hierarchical data structure to identify an allocator page identifying a region of a backing store having a greatest number of free blocks, and allocating the one or more blocks.

Abstract: Various embodiments set forth techniques for free space management in a block store. The techniques include receiving a request to allocate one or more blocks in a block store, accessing a sparse hierarchical data structure to identify an allocator page identifying a region of a backing store having a greatest number of free blocks, and allocating the one or more blocks.