Abstract: In a distributed data processing system with a set of multiple nodes, a first data shuffle memory pool is maintained at a data shuffle writer node, and a second data shuffle memory pool is maintained at a data shuffle reader node. The data shuffle writer node and the data shuffle reader node are part of the set of multiple nodes of the distributed data processing system. In-memory compression is performed on at least a portion of a data set from the first data shuffle memory pool. At least a portion of the compressed data is transmitted from the first data shuffle memory pool to the second data shuffle memory pool in a peer-to-peer manner. Each of the first data shuffle memory pool and the second data shuffle memory pool may include a hybrid memory configuration.

Abstract: Systems, methods, and articles of manufacture comprising processor-readable storage media are provided to implement read-ahead memory operations using learned memory access patterns for memory management systems. For example, a method for managing memory includes receiving a request from requestor (e.g., an active process) to perform a memory access operation, which includes a requested memory address. A determination is made as to whether a data block (e.g., page) associated with the requested memory address resides in a cache memory.

Abstract: Systems, methods, and articles of manufacture comprising processor-readable storage media are provided for implementing an accelerated data storage synchronization to protect data in a storage system. For example, a first server node accesses a data block from a protected storage system in response to a request from an application executing on the first server node. A modified data block is written to a write-back cache of the first server node. A synchronization operation is performed in response to a request by the application, which includes sending a copy of the modified data block in the write-back cache to a second server node to temporarily store the copy of the modified data block in a storage media of the second server node, and informing the application that the synchronization operation is deemed complete after the copy of the modified data block is stored in the storage media of the second server node.

Abstract: A data element to be inserted into a memory data structure, represented by a key and a value, is received. A target node into which the received data element is to be inserted is determined based on the key of the received data element. A determination is made whether or not the target node is already compressed. An append-write operation to insert the data element into the target node is performed when the target node is already compressed. An evaluation is performed prior to inserting the data element when the target node is not already compressed. An in-place write operation to insert the data element into the uncompressed target node is performed when the evaluation generates a first result. The target node is compressed and then an append-write operation to insert the data element into the compressed target node is performed when the evaluation generates a second result.

Abstract: Systems and methods are provided for implementing accelerated data access operations. For example, a method for accessing data includes loading an image into a memory, wherein the image comprises a plurality of image layers, and wherein each image layer references one or more data blocks. A request is received for a target data block from an active process operating on the image. Data structures associated with the target data bock are then accessed, wherein the data structures comprise metadata of a previous lookup path that was performed across one or more of the image layers of the image to access the target data block. The metadata of the data structures is utilized to access an existing in-memory copy of the target data block without having to traverse the previous lookup path, and the in-memory copy of target data block is returned to the active process.

Abstract: Data backup systems and methods are provided to asynchronously mirror in-memory checkpoint images generated by a process executing on a host system. For example, a data backup system is configured to: receive a checkpoint request message to store a copy of a checkpoint image which includes memory pages residing in a first memory region of memory on the host system; pull a copy of metadata associated with the memory pages from a second memory region of memory on the host system; pull a copy of the memory pages of the checkpoint image from the first memory region; generate a copy of the checkpoint image using the memory pages and associated metadata pulled from the host system; and store the copy of the checkpoint image in a storage device with an in-memory format that is the same as the checkpoint image residing in the first memory region on the host system.

Abstract: At least one memory region from a memory device of a host computing system is identified as containing data considered to be inactive. The host computing system is operatively coupled to a storage system external to the host computing system and configured to execute an application program. At least one memory region is stored as an archived memory region on a non-volatile storage device of the storage system in a memory format utilized on the host computing system.

Abstract: Systems, methods, and articles of manufacture comprising processor-readable storage media are provided for managing memory by persistence of page access heuristics. For example, a method includes collecting, during run-time execution of an application, access heuristics with regard to each page of data that is accessed by the application and cached in a page cache associated with the application, and generating metadata for each cached page in the page cache. The metadata for a given cached page represents the access heuristics of the given cached page. The metadata for each cached page is maintained in a volatile memory during run-time execution of the application. Moreover, the metadata for each cached page is persistently stored in a non-volatile memory during run-time execution of the application. The persistently stored metadata for each cached page is accessed when the application is restarted to restore the page cache associated with the application to a previous state.