Abstract: Techniques for managing computing resources are disclosed. In one particular embodiment, the techniques may be realized as a method for managing computing resources including receiving a desired resource configuration, determining currently available resources, determining, via at least one computer processor, whether the currently available resources satisfy the desired resource configuration, and determining a resource implementation based on the determination of whether the currently available resources satisfy the desired resource configuration.

Abstract: Information cluster systems and methods are presented. In one embodiment, a cluster method comprises: performing an engine process including issuing requests to bring a resource online, offline, and monitor the resources, wherein the engine process is performed by an engine; performing a resource interaction process including interacting with a resource and directing a resource to comply with the request from the engine process, wherein the resource interaction process is performed by a resource interaction agent; performing a predicate logic process including performing predicate logic operations to determine if a predicate logic condition associated with the resource is satisfied and forwarding an indication of the results of the predicate logic operations to the engine process, wherein the predicate logic process is performed by a predicate logic agent that is separate from the engine performing the engine process.

Abstract: A computer-implemented method for managing sub-clusters within dependent clustered computing systems subsequent to partition events may include (1) identifying a first computing cluster and a second computing cluster, (2) determining that the first computing cluster depends on the second computing cluster, (3) detecting a partition event that partitions the first computing cluster into multiple sub-clusters, (4) determining, in response to detecting the partition event, whether the second computing cluster is partitioned into multiple sub-clusters, and (5) making, based at least in part on determining whether the second computing cluster is partitioned into multiple sub-clusters, a fencing decision for the multiple sub-clusters of the first computing cluster. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A computer-implemented method for changing fencing modes in clusters may include (1) identifying a cluster of nodes configured with a fencing mode for resolving split-brain scenarios within the cluster, (2) identifying a request to reconfigure the cluster of nodes with a new fencing mode, (3) transmitting the request to each node in the cluster of nodes, (4) in response to the request, for each node in the cluster of nodes, testing the operability of the new fencing mode from the node, (5) receiving a message from the node indicating the success of the test; and, after receiving the message indicating the success of the test from every node within the cluster of nodes, and (6) reconfiguring each node within the cluster of nodes to use the new fencing mode. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The failure of a specific link of a specific node of a cluster is detected, for example through the operating system of the specific node. A notification of the failure of the specific link is disseminated to the nodes of the cluster, such that the nodes receive the notification prior to learning of the link-down event from the expiration of corresponding heartbeats, and process the event in parallel. The notification of the link failure can be disseminated by broadcasting a message notifying the nodes of the failure of the specific link. The notification of the link failure can also be disseminated by transmitting the notification to a centralized computing device which is polled by the nodes of the cluster at a specific frequency for notifications of link failures.

Abstract: The failure of a specific link of a specific node of a cluster is detected, for example through the operating system of the specific node. A notification of the failure of the specific link is disseminated to the nodes of the cluster, such that the nodes receive the notification prior to learning of the link-down event from the expiration of corresponding heartbeats, and process the event in parallel. The notification of the link failure can be disseminated by broadcasting a message notifying the nodes of the failure of the specific link. The notification of the link failure can also be disseminated by transmitting the notification to a centralized computing device which is polled by the nodes of the cluster at a specific frequency for notifications of link failures.

Abstract: Information cluster systems and methods are presented. In one embodiment, a cluster method comprises: performing an engine process including issuing requests to bring a resource online, offline, and monitor the resources, wherein the engine process is performed by an engine; performing a resource interaction process including interacting with a resource and directing a resource to comply with the request from the engine process, wherein the resource interaction process is performed by a resource interaction agent; performing a predicate logic process including performing predicate logic operations to determine if a predicate logic condition associated with the resource is satisfied and forwarding an indication of the results of the predicate logic operations to the engine process, wherein the predicate logic process is performed by a predicate logic agent that is separate from the engine performing the engine process.

Abstract: A computer-implemented method for managing sub-clusters within a multi-cluster computing system subsequent to a network-partition event may include: 1) identifying a plurality of computer clusters, 2) identifying a network-partition event that partitions each of the plurality of computer clusters into multiple sub-clusters, 3) identifying, within the multiple sub-clusters, a plurality of nodes that share a common attribute or property, and then 4) ensuring that only those nodes that share the common attribute survive subsequent to the network-partition event. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A method involves: executing an application on a cluster, which includes several nodes; and, subsequent to the cluster being partitioned into several sub-clusters, a first one of the sub-clusters executing a first one of the applications and a second one of the sub-clusters being capable of executing a second application. The first sub-cluster executes the first one of the applications after either determining that no other sub-cluster includes nodes that are part of the logical cluster that runs that application or after winning ownership of the logical cluster that runs the application. In a competition for ownership of a logical cluster, the winning sub-cluster is selected based on application-specific information, which is specific to the application that runs on that logical cluster.

Abstract: A method or apparatus for cooperative data replication. The method in one embodiment can be performed by a computer system or several computer systems executing software instructions. The method may include modifying data in n data blocks of a data volume to create n modified data blocks. A copy of each of the n modified data blocks is created for subsequent transfer to other nodes. A first computer system transmits the n modified data block copies to n nodes, respectively, wherein each of the n nodes comprises a second computer system and a replica of the data volume prior to the modification of data in the n data blocks. Thereafter, one of the n second computer systems transmits a copy of the modified data block copy it receives to another of the n second computer systems.