Abstract: The disclosure relates to technology for predicting file access patterns by identifying one or more files stored in a first cache and one or more pages associated with each of the one or more files in a second cache. The one or more files in the first cache are evicted based on a first eviction policy and, for each of the one or more files, the one or more pages in the second cache are evicted based on a second eviction policy. Access patterns of the one or more files based on the first and second eviction policies may then be predicted.

Abstract: The disclosure relates to technology for predicting file access patterns by identifying one or more files stored in a first cache and one or more pages associated with each of the one or more files in a second cache. The one or more files in the first cache are evicted based on a first eviction policy and, for each of the one or more files, the one or more pages in the second cache are evicted based on a second eviction policy. Access patterns of the one or more files based on the first and second eviction policies may then be predicted.

Abstract: An approach to offloading probabilistic computations is described. An application server comprising a memory and a processor and coupled to a network-attached storage device configured to create a dedicated process in response to a procedural call to a virtual machine container based on a data request is disclosed. The processor forwards the data request to the network-attached storage device, programs one or more virtual machines to perform a probabilistic computation based on the procedural call, and directs the probabilistic computation to a first virtual machine of the one or more virtual machines. The request for data is transformed into a modified call using a virtualized lookup call.

Abstract: An approach to offloading probabilistic computations is described. An application server comprising a memory and a processor and coupled to a network-attached storage device configured to create a dedicated process in response to a procedural call to a virtual machine container based on a data request is disclosed. The processor forwards the data request to the network-attached storage device, programs one or more virtual machines to perform a probabilistic computation based on the procedural call, and directs the probabilistic computation to a first virtual machine of the one or more virtual machines. The request for data is transformed into a modified call using a virtualized lookup call.

Abstract: An instruction from an application server to perform a computation is received at a network-attached storage (NAS) device. The computation uses arrays of data that are stored by the NAS device as inputs. The instruction includes remote procedure calls that identify operations that are included in the computation, including a first remote procedure call that will cause the NAS device to perform a read operation on a first file containing an array of data to be used as an input for the computation, and a second remote procedure call that will cause the NAS device to perform an array operation using the array of data. The operations are executed on the NAS device to produce a result that is stored in a second file in a location in a file system managed by the NAS device and accessible to the application server.

Abstract: An apparatus is configured to perform a method for a graph based computation. The method includes receiving a procedural call from an application server, the procedural call comprising at least one primitive extended from a distributed file system (DFS) protocol. The method also includes initiating at least one virtual machine. The method further includes performing a graph based computation based on the procedural call using the at least one virtual machine. The method still further includes transmitting a result of the graph based computation to the application server.

Abstract: An instruction from an application server to perform a computation is received at a network-attached storage (NAS) device. The computation uses arrays of data that are stored by the NAS device as inputs. The instruction includes remote procedure calls that identify operations that are included in the computation, including a first remote procedure call that will cause the NAS device to perform a read operation on a first file containing an array of data to be used as an input for the computation, and a second remote procedure call that will cause the NAS device to perform an array operation using the array of data. The operations are executed on the NAS device to produce a result that is stored in a second file in a location in a file system managed by the NAS device and accessible to the application server.

Abstract: An apparatus comprising an input/output (IO) port configured to couple to a large-scale storage device, a memory configured to store a plurality metadata databases (DBs) for a file system of the large-scale storage device, wherein the plurality of metadata DBs comprise key-value pairs with empty values, and a processor coupled to the IO port and the memory, wherein the processor is configured to partition the file system into a plurality of partitions by grouping directories in the file system by a temporal order, and index the file system by storing metadata of different partitions as keys in separate metadata DBs.

Abstract: A method implemented on a node, the method comprising receiving a key according to a sub-index of the key, wherein the sub-index identifies the node, and wherein the key corresponds to a data segment of a file, determining whether the data segment is stored in a data storage system according to whether the key appears in a hash table.