Abstract: A technique monitors input/output (I/O) and Persistent Reservation (PR) activity patterns to detect degraded performance of a highly available and fault tolerant application executing in a multi-site disaster recovery (DR) environment. Multiple instances of the application execute in different virtual machines (VMs) of a compute layer within a guest clustering configuration that extends across clusters of the sites. A storage layer of the clusters provides shared storage to the multiple VMs across the multiple sites. One of the sites is configured as an active storage site configured to receive and service I/O requests from the compute layer. A single instance of the application is active at a time and configured as a “compute owner” of the shared storage to issue the I/O requests to the shared storage. The compute owner and active storage site may not be co-located on the same site, leading to excessive I/O and PR activity patterns indicative of degraded performance.

Abstract: In accordance with some aspects of the present disclosure, a non-transitory computer readable medium is disclosed. In some embodiments, the non-transitory computer readable medium includes instructions that, when executed by a processor, cause the processor to receive, from a workload hosted on a host of a cluster, first I/O traffic programmed according to a first I/O traffic protocol supported by a cluster-wide storage fabric exposed to the workload as being hosted on the same host. In some embodiments, the workload is recovered by a hypervisor hosted on the same host. In some embodiments, the non-transitory computer readable medium includes the instructions that, when executed by the processor, cause the processor to adapt the first I/O traffic to generate second I/O traffic programmed according to a second I/O traffic protocol supported by a repository external to the storage fabric and forward the second I/O traffic to the repository.

Abstract: Various embodiments set forth techniques for managing metadata for a vblock include dynamically normalizing and denormalizing vblock metadata associated with an extent. Vblock metadata associated with an extent is normalized when the extent is migrated to a different extent group by having the vblock metadata to a mapping between the extent identifier and an extent group identifier in a metadata map separate from the vblock metadata. Vblock metadata associated with an extent is denormalized whenever the number of vblock metadata associated with an extent drops below a threshold. Vblock metadata is denormalized by updating the vblock metadata to include a mapping to an extent group, based on a mapping of the extent to the extent group in a separate metadata map, and then removing the mapping in the separate metadata map.

Abstract: In accordance with some aspects of the present disclosure, a non-transitory computer readable medium is disclosed. In some embodiments, the non-transitory computer readable medium includes instructions that, when executed by a processor, cause the processor to receive, from a workload hosted on a host of a cluster, first I/O traffic programmed according to a first I/O traffic protocol supported by a cluster-wide storage fabric exposed to the workload as being hosted on the same host. In some embodiments, the workload is recovered by a hypervisor hosted on the same host. In some embodiments, the non-transitory computer readable medium includes the instructions that, when executed by the processor, cause the processor to adapt the first I/O traffic to generate second I/O traffic programmed according to a second I/O traffic protocol supported by a repository external to the storage fabric and forward the second I/O traffic to the repository.

Abstract: A high frequency snapshot technique improves data replication in a disaster recovery (DR) environment. A base snapshot is generated from failover data at a primary site and replicated to a placeholder file at a secondary site. Upon commencement of the base snapshot generation and replication, incremental light weight snapshots (LWSs) of the failover data are captured and replicated to the secondary site. A staging file at the secondary site accumulates the replicated LWSs (“high-frequency snapshots”). The staging file is populated with the LWSs in parallel with the replication of the base snapshot at the placeholder file. At a subsequent predetermined time interval, the accumulated LWSs are synthesized to capture a “checkpoint” snapshot by applying and pruning the accumulated LWSs at the staging file. Once the base snapshot is fully replicated, the pruned LWSs are merged to the base snapshot to synchronize the replicated failover data.

Abstract: In accordance with some aspects of the present disclosure, a non-transitory computer readable medium is disclosed. In some embodiments, the non-transitory computer readable medium includes instructions that, when executed by a processor, cause the processor to receive, from a workload hosted on a host of a cluster, first I/O traffic programmed according to a first I/O traffic protocol supported by a cluster-wide storage fabric exposed to the workload as being hosted on the same host. In some embodiments, the workload is recovered by a hypervisor hosted on the same host. In some embodiments, the non-transitory computer readable medium includes the instructions that, when executed by the processor, cause the processor to adapt the first I/O traffic to generate second I/O traffic programmed according to a second I/O traffic protocol supported by a repository external to the storage fabric and forward the second I/O traffic to the repository.

Abstract: The present disclosure provides a method, apparatus, and computer program product for generating consistent snapshots without quiescing applications. The snapshots correspond to consistency groups that in turn correspond to one or more vDisks on one or more nodes. Some embodiments of the present disclosure address the problem of generating consistent backup snapshots that cover cases of multiple virtual machines (VMs) that operate in a multi-node distributed computing environment (e.g., in clusters) without quiescing the accessing entity(ies). In some embodiments, the process includes identification of a triggering snapshot event, disabling of write complete responses to requesting entities for the resources corresponding to the snapshot to be generated, generating snapshots while one or more additional write requests are received, and using one or more techniques to insure that all actions corresponding to the snapshotting function do not occur outside of a specified time frame.

Abstract: Systems for multi-cluster virtualized computing system management. A method for performing virtual entity replication between source computing clusters and target computing clusters commences upon establishing a virtual entity naming convention that is observed by both the source computing clusters and the target computing clusters. A snapshot from a source cluster is associated with a global snapshot ID before being transmitted to a target computing cluster. At some point in time, the source cluster will initiate acts to replicate a virtual entity to a particular data state that is associated with a particular named snapshot. A second replication protocol then commences. The second replication protocol includes exchanges that serve to determine whether or not the target computing cluster has a copy of a particular named snapshot as named by the global snapshot ID, and if so, to then initiate virtual entity replication at the target computing cluster using the named snapshot.

Abstract: Systems for multi-cluster virtualized computing system management. A method for performing virtual entity replication between source computing clusters and target computing clusters commences upon establishing a virtual entity naming convention that is observed by both the source computing clusters and the target computing clusters. A snapshot from a source cluster is associated with a global snapshot ID before being transmitted to a target computing cluster. At some point in time, the source cluster will initiate acts to replicate a virtual entity to a particular data state that is associated with a particular named snapshot. A second replication protocol then commences. The second replication protocol includes exchanges that serve to determine whether or not the target computing cluster has a copy of a particular named snapshot as named by the global snapshot ID, and if so, to then initiate virtual entity replication at the target computing cluster using the named snapshot.

Abstract: The present disclosure provides a method, apparatus, and computer program product for generating consistent snapshots without quiescing applications. The snapshots correspond to consistency groups that in turn correspond to one or more vDisks on one or more nodes. Some embodiments of the present disclosure address the problem of generating consistent backup snapshots that cover cases of multiple virtual machines (VMs) that operate in a multi-node distributed computing environment (e.g., in clusters) without quiescing the accessing entity(ies). In some embodiments, the process includes identification of a triggering snapshot event, disabling of write complete responses to requesting entities for the resources corresponding to the snapshot to be generated, generating snapshots while one or more additional write requests are received, and using one or more techniques to insure that all actions corresponding to the snapshotting function do not occur outside of a specified time frame.

Abstract: A method and apparatus for providing dynamic multi-pathing for an asymmetrical logical unit access (ALUA) based storage system. The method comprises identifying a first processor within a storage system as providing an optimized path to a disk array, identifying a second processor within a storage system as providing an unoptimized path to the disk array and, in response to failure of the optimized path, immediately switching from the optimized path to the unoptimized path.