Abstract: A method for managing tasks in a storage system, the method may include: (a) obtaining, by a scheduler, a shared budget for background storage tasks and foreground storage tasks; (b) obtaining, by the scheduler, a background budget for background storage tasks; wherein the background budget is a fraction of the shared budget; (c) allocating, by the scheduler, resources to pending storage tasks according to the shared budget and the background budget; wherein the allocating comprises (i) allocating the shared budget while prioritizing foreground storage tasks over background storage tasks; and (ii) allocating the background budget to background storage tasks; and (d) participating, by the scheduler, in executing of storage tasks according to the allocation.

Abstract: A method for accessing a database that is stored in a storage system, the method includes: (a) receiving a request to obtain a database entity (DE) of a requested version line out of multiple version lines that differ from each other by one or more transactions; (b) obtaining, for each DE portion out of multiple DE portions of the DE, content of the DE portion as written during a transaction of the requested version line that was a last transaction of the requested version line that wrote the DE portion to the database; wherein the obtaining includes (i) using first metadata that is indicative of transactions of the requested version line, and (ii) using second metadata that is indicative of which transactions were involved in writing to a logical address range that is associated with the DE portion; and (d) sending a response to the request, wherein the response comprises, for each DE portion of the DE portions, the obtained content.

Abstract: A method for obtaining information related to a filesystem path in a file system that is stored in a storage system, the method may include (i) receiving a request to obtain hierarchy information related to file system entities stored under the filesystem path; and (ii) retrieving the hierarchy information; wherein the retrieving comprises accessing a filesystem hierarchy table that stores information regarding pathnames of file system entities of the file system; wherein the retrieving includes searching for path components of the filesystem path in subsets of cells of the filesystem hierarchy table, wherein each subset of cells is part of a set of cells that include path components of a certain pathname of a filesystem entity that belong to the file system.

Abstract: A method for network group based access control, the method may include (i) receiving a request to access content stored in a storage system, the request is associated with a user and a network group that is suggested to include the user; (ii) generating a user key based on an identifier of the user and the network group; (iii) determining, based in part on the user key, whether the user is a current user of the network group, and (iv) declining the request when the user is not the current user of the network group. The determining may include determining that the user is not a current user of the network group when a user finding timestamp of the user precedes a network group finding timestamp of the network group.

Abstract: A method for protecting a disaster recovery site, the method may include receiving by source compute nodes of a storage system, during source storage periods, write requests for storing content in the storage system; writing by source compute nodes, during the source storage periods, the content into the storage nodes of the storage system; maintaining replication compute nodes of the storage system deactivated during the source storage periods; reading the content by the replication compute nodes from the storage nodes during replication periods; participating, by the replication compute nodes, in outputting the content to one or more data recovery sites during the replication periods; and maintaining the source compute nodes deactivated during the source storage periods.

Abstract: A method for protecting a lock data structure, the method includes obtaining, by a compute entity of a storage system, a lock recovery indication and a cycle counter; executing, by the compute entity, at least a part of the current lock recovery process, wherein the executing of the at least part of the current lock recovery process includes (i) determining, by the compute entity, using a compute entity data structure that is indicative of lock fields that represent locks acquired by the compute entity, whether the segment comprised a compute-entity-relevant lock field, wherein the compute-entity-relevant lock field (a) belonged to the segment upon the occurrence of the failure and (b) represents a lock was acquired by the compute entity; and (ii) re-acquiring, by the compute entity, a re-acquired lock that corresponds to the compute-entity-relevant lock field, when it is determined that the segment comprised the compute-entity-relevant lock field.

Abstract: Techniques for performing large tensor tiling (LTT) in hardware are enabled. LTT divides a large tensor (e.g., of unsupported size) into overlapping tiles (e.g., having supported tensor size(s)). A tensor may be processed processing the tiles. The output of each processed tile is stored, for example, in a systolic array considering the tile's placement in the large tensor. The output of all processed tiles is identical to the output of processing the large tensor. Tiles may be processed by reusing data overlapping boundaries shared with other tiles. In some examples, overlapping data may be reused (e.g., written once) or partly reused (e.g., written twice). Tiling large tensors with boundary duplication supports dynamic adaptation to a wide variety of tensor sizes, avoids re-reading duplicated data, and avoids reorganizing hardware for large tiles, which reduces power consumption and area, reduces complexity, and increases processing efficiency.

Abstract: A neural processing unit (“NPU”) is enabled to process tensor format and dimensional changes within the NPU during import and export of tensor data to and from the NPU, and between execution of successive computation steps (e.g., between successive convolution operations being executed in sequence). The NPU features an input data handler, an N×M systolic array and an output data handler. Such in-place tensor format changes are enabled by equipping an input data handler with format change hardware, and operating the NPU such that the input data handler and output data handler and related hardware tracks the format state of tensors moving into and out of an N×M systolic array and may therefore alter the format of such tensors on the fly.

Abstract: A method for managing a transaction data structure, the method may include (a) receiving, at a first point in time that corresponds to a first value of a clock signal, a request to commit a transaction, and updating a status of the transaction, within the transaction data structure, as being a commit in progress; (b) updating the status of the transaction within the transaction data structure as being committed following (i) a completion of the committing, and (ii) a change of a value of the clock signal from the first value to a second value that differs from the first value; and (c) updating, using a background process, a database segment (DS) version metadata associated with the transaction regarding the completion of the committing. The DS version metadata belongs to a DS version metadata data structure. The DS version metadata data structure and the transaction data structure are selectively accessed during transactions.

Abstract: Techniques for dynamic concatenation of CNN tensor space in hardware are enabled. A concatenation operation may be performed by hardware-implemented data routing that routes data into a systolic array data structure. Tensor channels may be distributed over the systolic array to implement the concatenation without overhead or software. Technical advantages include reduced CPU operations, reduced access to SRAM, reduced power consumption, faster tensor operations, etc. For example, a computing system with an NPU may include a systolic array of PEs with data memories. A data router determines tensor concatenation routing to the PE data memories based on the size and number of tensors in tensor packages (e.g., a first tensor package with m tensors and a second tensor package comprising n tensors) may be routed for storage in PE data memories. The m and n stored tensors are concatenated in the systolic array.

Abstract: A method for storing a received data chunk (DC) in a storage system, the method includes (a) obtaining a received fingerprint of the received DC, the received fingerprint may include received fingerprint elements that are indicative of occurrences, within the received DC, of content elements, the received fingerprint elements are ordered according to a given order; (b) searching, within a tree, for a similar stored fingerprint; the tree may include tree nodes that represent multiple stored fingerprints of stored data chunks that are stored in the storage system; different levels of the tree are allocated to different content elements; (c) compressing, when finding the similar stored fingerprint, the received DC based on a similar DC associated with the similar stored fingerprint, and updating storage system metadata to indicate that the received DC is stored in the storage system in a compressed form, and based on the similar stored DC.

Abstract: A method for managing access to a shared file system entity (FSE), the method includes (i) managing access requests for reading and writing accessed parts of the FSE, by a current point of access storage system (POASS), the current POASS currently acts as a point of access (POA) for the access requests; wherein stored parts of the FSE are stored at multiple POASSs, the multiple POASSs comprise the current POASS and former POASSs; (ii) determining, by the current POASS and while managing the access requests, to store the stored parts of the FSE at a fewer number of the former POASSs; and (iii) reducing, by two or more of the former POASSs, the number of former POASSs that store the stored parts of the FSE.

Abstract: Methods, systems are provided for reconfiguring the position of a first tap in a descrambler circuit LFSR after the LFSR has been trained and synchronized with a corresponding scrambler circuit LFSR. A data path from the second tap position to the descrambler output by-passes logic elements located in the data path from the first tap to the descrambler output, thereby reducing delay in the descrambler circuit after the reconfiguration (i.e., the “lock-shift” operation). The tap position change may be communicated by a mode manager to a corresponding scrambler circuit, for applying a matching reconfiguration in the scrambler circuit, either directly via an I/O line or indirectly. The indirect route includes in-band transmissions between two ICs with two sets of self-synchronizing scrambler/descrambler pairs, and is based on monitored receiver LFSR output signals that indicate when a scrambler/descrambler pair is synchronized or whether the output of a descrambler circuit comprises descrambled data.

Abstract: A method for receiving paths information of a file system (FS) that is stored in a storage system, the method includes (a) receiving a request to receive paths information of at least part of the FS; wherein the request is received at a request point in time (PIT); (b) determining a current path name information of the at least part of the FS, wherein the determining is based on: (i) a FS paths data structure that includes path name information that is updated up to a last update PIT, and (ii) FS paths change information that is indicative of FS path names changes introduced between the last update PIT and the request PIT; wherein the FS paths change information associates at least one of excluded FSEs and included FSEs for each direct or indirect ancestor FSE of any FSE that had its path name changed between the last update PIT and the request PIT; and (c) generating a response to the request, wherein the generating of the response comprises using the current FS path name information of the FS.

Abstract: A method for compression of cells of a tabular database (TD), the method may include (i) receiving data elements, by a storage system; (ii) calculating a compressed representation (CR) of the data elements by applying a compression process that includes (ii.a) defining sets based on data element values, (ii.b) associating a set value per set; and (ii.c) generating the CR of the data elements based on differences between the data elements and corresponding set values; (iii) generating compression metadata related to the CR of the data elements, wherein the compression metadata comprises the set values and a mapping between the CR of the data elements and the set values; and (iv) storing, in the TD, the CR of the data elements.

Abstract: A method for importing data to a tabular database (TD), the method includes: determining that data was written, using a file system command, to a file system location; wherein the file system location is associated with at least a part of the TD; and automatically importing the data to the at least part of the TD.

Abstract: A method for obtaining information related to a filesystem path in a file system that is stored in a storage system, the method may include (i) receiving a request to obtain hierarchy information related to file system entities stored under the filesystem path; and (ii) retrieving the hierarchy information; wherein the retrieving comprises accessing a filesystem hierarchy table that stores information regarding pathnames of file system entities of the file system; wherein the retrieving includes searching for path components of the filesystem path in subsets of cells of the filesystem hierarchy table, wherein each subset of cells is part of a set of cells that include path components of a certain pathname of a filesystem entity that belong to the file system.

Abstract: A method and system for characterization of Cannabaceae plants using macro photography images is disclosed. The method comprises the steps of receiving one or more macro photography images of a Cannabaceae plant; performing feature extraction analysis of trichomes using image processing, and performing plant characterization analysis using a neural network which analyzes the macro photography images. The training phase of the neural network comprises using results of chemical composition laboratory tests performed on the plants for which the macro photography images have been used in the training phase. The invention calculates and reports an assessment of maturity of the plant for harvesting, diagnosis of the existence of diseases, insects, or pests, assessment of the presence and concentrations of central ingredients, recommendations for treatment during plants drying, curing or storage production processes, and assessment of the quality and pricing of Cannabaceae plants products.

Abstract: Methods, systems are provided for reconfiguring the position of a first tap in a descrambler circuit LFSR after the LFSR has been trained and synchronized with a corresponding scrambler circuit LFSR. A data path from the second tap position to the descrambler output by-passes logic elements located in the data path from the first tap to the descrambler output, thereby reducing delay in the descrambler circuit after the reconfiguration (i.e., the “lock-shift” operation). The tap position change may be communicated by a mode manager to a corresponding scrambler circuit, for applying a matching reconfiguration in the scrambler circuit, either directly via an I/O line or indirectly. The indirect route includes in-band transmissions between two ICs with two sets of self-synchronizing scrambler/descrambler pairs, and is based on monitored receiver LFSR output signals that indicate when a scrambler/descrambler pair is synchronized or whether the output of a descrambler circuit comprises descrambled data.

Abstract: A method for managing a read statement of a transaction, the method may include: (a) receiving the read statement, wherein the read statement is for receiving a data unit that is stored in a relational database; (b) finding one or more candidate versions of the data unit; wherein the finding may include: (i) accessing a database segment (DS) version metadata of one or more DS versions of one or more DSs that are related to the data unit and are stored in the relational database; (ii) performing a timestamp based determination that may include checking whether to ignore a database segment version when the database segment version metadata identifies the database segment version by a database segment commit timestamp; and (iii) performing a transaction metadata based determination of whether to ignore the database segment version when the database segment version metadata identifies the database segment version by a transaction identifier; and (c) responding to the read statement by outputting a version of the