Abstract: A computer-implemented method for shipping I/O operations to prevent replication failure may include 1) attempting to perform an I/O operation in a system configured to replicate data from a data cluster to another data cluster, 2) detecting a failure in at least part of the attempt to perform the I/O operation that threatens to fail the system's replication of data from the data cluster to the other data cluster, and, in response to detecting the failure, 3) shipping the I/O operation from a node originally responsible for servicing the I/O operation to another node to complete the I/O operation without failing the system's replication of data from the data cluster to the other data cluster. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A computer-implemented method for shipping I/O operations to prevent replication failure may include 1) attempting to perform an I/O operation in a system configured to replicate data from a data cluster to another data cluster, 2) detecting a failure in at least part of the attempt to perform the I/O operation that threatens to fail the system's replication of data from the data cluster to the other data cluster, and, in response to detecting the failure, 3) shipping the I/O operation from a node originally responsible for servicing the I/O operation to another node to complete the I/O operation without failing the system's replication of data from the data cluster to the other data cluster. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A method and system for load balancing. The method includes discovering each of a plurality of hosts in a cluster, wherein the plurality of hosts is configured for accessing a LUN of a storage system through a storage network fabric. Global input/output (I/O) load characteristics are collected for each of the plurality of hosts at the device and/or volume level. A selected host is determined for processing an I/O originating at the local host, wherein the host is selected based on a current set of the global I/O load characteristics.

Abstract: A list of servers known to a client is compared with lists of servers stored on the servers. An instance of content (e.g., a data file) can be accessed from a server on the list if at least a simple majority of the servers have the same version of the list.

Abstract: A preferred node is selected for a specific functional role in a cluster. Dynamic and static parameters concerning the nodes are measured, including connectivity to shared storage. A user preference value is gleaned for each node, quantifying the user's preference of each corresponding node for the functional role. A preference rating is calculated for each node, based on the measured parameters and the gleaned user preference value. The preference rating indicates the node's suitability for the specific functional role, relative to the other nodes. Connectivity to shared storage in the cluster can be weighted more heavily than other parameters. Examples of specific functional roles in the cluster include becoming the master node, becoming a failover target for a specific application, or remaining in operation with access to shared cluster storage, in response to an occurrence of split brain.

Abstract: A computer-implemented method for transferring input/output operations within computer clusters may include (1) identifying a computer cluster that includes (i) a plurality of nodes configured to facilitate access to a plurality of physical storage devices and (ii) at least one computing object that represents data stored on the physical storage devices, (2) detecting at least one failure that prevents at least one node within the computer cluster from performing at least one input/output operation on the computing object that represents the data stored on the physical storage devices, (3) identifying at least one other node within the computer cluster that is able to perform the input/output operation on the computing object, and then (4) directing the other node within the computer cluster to perform the input/output operation on the computing object. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The tracking of cluster-wide connectivity information is optimized. I/O errors that result from failed operations in which specific nodes attempt to access shared storage are detected. The start-times of failed operations are tracked. The tracked start-times are compared to the time at which the cluster-wide connectivity information was last updated. Responsive to the results of the comparing, the cluster-wide connectivity information is updated in response only to a single I/O error that results from a single failed operation that was initiated after the update time, wherein additional errors resulting from failed operations with start-times after the update time are also detected, thereby preventing redundant updates of the cluster-wide connectivity information. The update time is set to the time at which the cluster-wide connectivity information is updated.

Abstract: A method and system for load balancing. The method includes discovering each of a plurality of hosts in a cluster, wherein the plurality of hosts is configured for accessing a LUN of a storage system through a storage network fabric. Global input/output (I/O) load characteristics are collected for each of the plurality of hosts at the device and/or volume level. A selected host is determined for processing an I/O originating at the local host, wherein the host is selected based on a current set of the global I/O load characteristics.

Abstract: A node in a server cluster is designated as a quorum disk. The node stores a list of other nodes in the server cluster also designated as quorum disks. The node can replace the first list with a second and more recent list of quorum disks only if the second list is updated on at least a simple majority of quorum disks on the first list.

Abstract: A method and system for multi-staged recovery of a distributed computer system. The method includes receiving a failure event notification from at least one node of the distributed computer system and executing a plurality of recovery stages upon receiving the failure event notification by using a recovery manager, wherein each of the plurality of recovery stages performs a defined recovery task. The progress of recovery is tracked by using at least one state machine executed by the recovery manager, wherein the state machine reflects progress of each of the recovery stages. The progress of recovery is monitored to a completion by using the state machine and the recovery manager.

Abstract: A node in a server cluster is designated as a quorum disk. The node stores a list of other nodes in the server cluster also designated as quorum disks. The node can replace the first list with a second and more recent list of quorum disks only if the second list is updated on at least a simple majority of quorum disks on the first list.

Abstract: A list of servers known to a client is compared with lists of servers stored on the servers. An instance of content (e.g., a data file) can be accessed from a server on the list if at least a simple majority of the servers have the same version of the list.

Abstract: A method and system for multi-staged recovery of a distributed computer system. The method includes receiving a failure event notification from at least one node of the distributed computer system and executing a plurality of recovery stages upon receiving the failure event notification by using a recovery manager, wherein each of the plurality of recovery stages performs a defined recovery task. The progress of recovery is tracked by using at least one state machine executed by the recovery manager, wherein the state machine reflects progress of each of the recovery stages. The progress of recovery is monitored to a completion by using the state machine and the recovery manager.

Abstract: It has been discovered that systems, methods, apparatus and software can accommodate the addition or removal of available physical storage (storage “expansion”) in a storage virtualization environment while virtual storage devices remain online and generally available to users. When a change to the available storage has occurred, new storage device geometry and configuration information reflecting the change is obtained and/or calculated. This new information is used to update mapping information used by virtualization software and/or to update information used by an operating system to manage storage devices. Such updating occurs while some or all of the virtual storage devices associated with the physical storage devices remain available to users. In some cases, I/O operations targeting a physical device are held, in a manner transparent to a user, while updating occurs.