One or more aspects of the present disclosure relate to cache memory management. In embodiments, a global memory of a storage array into one or more cache partitions based on an anticipated activity of one or more input/output (IO) service level (SL) workload volumes can be dynamically partitioned.

A request to perform a write operation to write data at a memory sub-system is received. Responsive to the request to perform the write operation, the data is stored at a cache portion of cache memory of the memory sub-system. A duplicate copy of the data is stored at a write buffer portion of cache memory. An entry of the write buffer record is recorded that maps a location of the duplicate copy of the data stored at the write buffer portion to a location of the data stored at the cache portion of the cache memory. A memory operation is performed at the memory sub-system based at least in part on the write buffer record.

Techniques for providing horizontally scaled caching of versioned data are provided. In some aspects, the techniques described herein relate to a method including initializing a first version cache (VC) object based on a version of data stored in a data storage device; replicating the first VC to generate a second VC; receiving a write operation at the first VC; generating a delta for the write operation, the delta representing a change in the version of data; writing the delta to a persistent replication log, the persistent replication log storing an ordered set of deltas including the delta; writing data in the write operation to the data storage device; and applying the ordered set of deltas at the second VC to update data stored by the second VC.

One or more aspects of the present disclosure relate to cache memory management. In embodiments, a global memory of a storage array into one or more cache partitions based on an anticipated activity of one or more input/output (IO) service level (SL) workload volumes can be dynamically partitioned.

A request to perform a write operation to write data at a memory sub-system is received. Responsive to the request to perform the write operation, the data is stored at a cache portion of cache memory of the memory sub-system. A duplicate copy of the data is stored at a write buffer portion of cache memory. An entry of the write buffer record is recorded that maps a location of the duplicate copy of the data stored at the write buffer portion to a location of the data stored at the cache portion of the cache memory. A memory operation is performed at the memory sub-system based at least in part on the write buffer record.

A method for demoting data elements from a cache is disclosed. The method maintains a heterogeneous cache comprising a higher performance portion and a lower performance portion. The method maintains, within the lower performance portion, a ghost cache containing statistics for data elements that are currently contained in the heterogeneous cache, and data elements that have been demoted from the heterogeneous cache within a specified time interval. The method maintains, for the ghost cache, multiple LRU lists that designate an order in which data elements are demoted from the lower performance portion. The method utilizes the statistics to determine in which LRU lists the data elements are referenced. A corresponding system and computer program product are also disclosed.

A system shares I/O load between controllers in a high availability system. For writes, a controller determines based on one or more factors which controller will flush batches of data from write-back cache to better distribute the I/O burden. The determination occurs after the local storage controller caches the data, mirrors it, and confirms write complete to the host. Once it is determined which storage controller will flush the cache, the flush occurs and the corresponding metadata at a second layer of indirection is updated by that determined storage controller (whether or not it is identified as the owner of the corresponding volume to the host, while the volume owner updates metadata at a first layer of indirection). For a host read, the controller that owns the volume accesses the metadata from whichever controller(s) flushed the data previously and reads the data, regardless of which controller had performed the flush.

A method comprising: receiving, at a vector processor, a request to store data; performing, by the vector processor, one or more transforms on the data; and directly instructing, by the vector processor, one or more storage device to store the data; wherein performing one or more transforms on the data comprises: erasure encoding the data to generate n data fragments configured such that any k of the data fragments are usable to regenerate the data, where k is less than n; and wherein directly instructing one or more storage device to store the data comprises: directly instructing the one or more storage devices to store the plurality of data fragments.

Provided are a computer program product, system, and method for using mirror indicators to determine whether to mirror tracks in a data set in a primary volume mirrored to a secondary volume. A table is read. The table is maintained by a primary controller managing the primary volume that includes a mirror indicator for each of a plurality of tracks in at least one data set configured in the primary volume indicating whether a track is to be mirrored to the secondary volume. Record sets are read from a cache of the primary controller for the tracks in primary volume having the mirror indicators in the table indicating that the track is to be mirrored. The write data in the read record sets is applied to tracks in the secondary volume mirroring the tracks in the primary volume.

A data access system has host computers having front-end controllers nFE_SAN connected via a bus or network interconnect to back-end storage controllers nBE_SAN, and physical disk drives connected via network interconnect to the nBE_SANs to provide a distributed, high performance, policy based or dynamically reconfigurable, centrally managed, data storage acceleration system. The hardware and software architectural solutions eliminate BE_SAN controller bottlenecks and improve performance and scalability. In an embodiment, the nBE_SAN (BE_SAN) firmware recognize controller overload conditions, informs Distributed Resource Manager (DRM), and, based on the DRM provided optimal topology information, delegates part of its workload to additional controllers. The nFE_SAN firmware and additional hardware using functionally independent and redundant CPUs and memory that mitigate single points of failure and accelerates write performance. The nFE_SAN and FE_SAN controllers facilitate Converged I/O Interface by simultaneously supporting storage I/O and network traffic.