Abstract: A technique for transmitting data involves creating a shared hash set for sharing pattern data between a protocol layer and an input/output (IO) path layer, wherein the shared hash set stores pattern keys and pattern content for use in pattern detection. The technique further involves receiving, by the protocol layer, a pattern data block from a client, wherein the pattern data block comprises the pattern data and non-pattern data. The technique further involves transmitting, by the protocol layer, the pattern data block to the IO path layer. The technique further involves writing the pattern data block to a common block file system (CBFS) layer.

Abstract: Techniques for recovering data involve, in response to detecting that a data block of a storage system is attacked, determining one or more attacked data blocks, wherein snapshots in multiple snapshots of the storage system have snapshot replicate numbers. Such techniques further involve determining one or more data block replicate numbers of the one or more attacked data blocks. Such techniques further involve identifying the latest target snapshot in the multiple snapshots based on the snapshot replicate numbers of snapshots in the multiple snapshots and the one or more data block replicate numbers, wherein all data blocks in the target snapshot are not attacked. Such techniques further involve recovering data of the storage system based on the latest target snapshot.

Abstract: Techniques for compressing a data block involve adding a compression flag to a data block after the data block is decompressed by a protocol layer, the compression flag indicating that the data block is compressible or uncompressible. Such techniques further involve acquiring the compression flag of the data block. Such techniques further involve determining, in response to the compression flag of the data block indicating that the data block is compressible, to perform inline compression on the data block. Such techniques further involve preventing, in response to the compression flag of the data block indicating that the data block is uncompressible, the inline compression from being performed on the data block.

Abstract: Techniques for transmitting a data block involve receiving, from a client, a request for reading the data block. Such techniques further involve determining, at a data block layer, whether the data block needs to be compressed at a protocol layer based on a flag indicating whether the data block has been compressed and a storage size of the data block. Such techniques further involve transferring, in response to determining that the data block does not need to be compressed at the protocol layer, an indication that the data block does not need to be compressed at the protocol layer from the data block layer to the protocol layer. Such techniques further involve sending the data block to the client through the protocol layer without compressing the data block at the protocol layer.

Abstract: Backing up data is described. An example method includes acquiring data block deduplication information of source data and pieces of backup node deduplication information of backup nodes in a backup node set. Then, based on the data block deduplication information and the pieces, deduplication rates of the source data are determined relative to the backup nodes in the backup node set. In addition, based on the deduplication rates for the plurality of backup nodes, a target backup node is selected for backing up the source data from the backup node set. Data features of the to-be-backed-up source data are utilized to determine the deduplication rate of the source data relative to each backup node, thereby selecting the target backup node through the deduplication rates of the backup nodes. The space occupation of backup nodes can also be reduced when backing up the source data, thereby achieving better space utilization.

Abstract: Techniques for data processing involve: performing, according to a determination that a programmable circuit receives user-written data, data feature detection by the programmable circuit on a first part of the user-written data. Such techniques further involve: replacing, in response to the first part matching a predetermined data feature, the first part by the programmable circuit with a data representation corresponding to the predetermined data feature. In addition, such techniques involve: writing the data representation to a cache region of a storage system. Accordingly, such techniques can save cache resources of a storage system and can save CPU of the storage system, thus avoiding the latency due to data feature detection and improving the user experience.

Abstract: Backing up data is described. An example method includes acquiring data block deduplication information of source data and pieces of backup node deduplication information of backup nodes in a backup node set. Then, based on the data block deduplication information and the pieces, deduplication rates of the source data are determined relative to the backup nodes in the backup node set. In addition, based on the deduplication rates for the plurality of backup nodes, a target backup node is selected for backing up the source data from the backup node set. Data features of the to-be-backed-up source data are utilized to determine the deduplication rate of the source data relative to each backup node, thereby selecting the target backup node through the deduplication rates of the backup nodes. The space occupation of backup nodes can also be reduced when backing up the source data, thereby achieving better space utilization.

Abstract: Techniques for compressing a data block involve adding a compression flag to a data block after the data block is decompressed by a protocol layer, the compression flag indicating that the data block is compressible or uncompressible. Such techniques further involve acquiring the compression flag of the data block. Such techniques further involve determining, in response to the compression flag of the data block indicating that the data block is compressible, to perform inline compression on the data block. Such techniques further involve preventing, in response to the compression flag of the data block indicating that the data block is uncompressible, the inline compression from being performed on the data block.

Abstract: Techniques for data migration involve obtaining feature data of a plurality of logical storage units in a first storage device, the feature data comprising deduplication information of each of the plurality of logical storage units. Such techniques further involve determining, based on the feature data, a set of similarity values between the plurality of logical storage units. Such techniques further involve migrating a logical storage unit having a similarity value higher than a similarity threshold among the plurality of logical storage units from the first storage device to a second storage device different from the first storage device.

Abstract: Embodiments of the present disclosure provide a method, an electronic device, and a computer program product for detecting a network attack. The method includes determining a compression ratio of written data in a time window of a predetermined time length, where the compression ratio is a ratio of a data amount of the written data before being compressed to a data amount of the written data after being compressed. The method further includes determining a user side connection associated with the written data at least according to determining that the compression ratio is less than a threshold compression ratio. In addition, the method includes determining data from the user side connection as a network attack according to determining that files associated with the user side connection cannot pass a file integrity check. Embodiments of the present disclosure can significantly reduce the computational amount of network attack detection.

Abstract: Techniques perform data processing. Such techniques involve obtaining, by a first operating system, first data of a storage system from a data collector. Such techniques further involve synchronizing the first data from the first operating system to a second operating system, wherein an architecture of the second operating system is different from that of the first operating system. Such techniques further involve, in response to that the first operating system has a failure, obtaining, by the second operating system, second data of the storage system from the data collector. Such techniques further involve, in response to that the first operating system is recovered, synchronizing the second data from the second operating system to the first operating system. Accordingly, such techniques can minimize processing time, ensure high availability of data collection, save computing and memory resources of a storage system, and help improve user experience.

Abstract: A method of performing synchronous replication from a primary storage system apparatus (PSSA) to a secondary storage system apparatus (SSSA) is provided. The method includes (a) in response to write requests received by the PSSA, (i) calculating metadata changes by the PSSA for accommodating the write requests, (ii) generating, by the PSSA, metadata journal log entries that describe the metadata changes, and (iii) mirroring the metadata journal log entries from the PSSA to the SSSA; (b) regenerating the metadata changes by the SSSA based on the metadata journal log entries mirrored from the PSSA to the SSSA; and (c) writing the regenerated metadata changes to persistent storage of the SSSA. A method performed by the SSSA is also provided. An apparatus, system, and computer program product for performing similar methods are also provided.

Abstract: Systems and methods for storage system attack detection and response are described. In an illustrative, non-limiting embodiment, a method may include: obtaining a plurality of count features corresponding to a respective plurality of time points, where the count features include a plurality of count values for data blocks of a storage object; determining, for a first time point, a plurality of corresponding correlation coefficients between a first count feature for the first time point and other count features for other time points; determining a score for the first time point based, at least in part, on the plurality of corresponding correlation coefficients; determining that the score is less than a predetermined threshold; determining, based on the score being less than the predetermined threshold, that the storage object corresponding to the first time point is under an attack; and providing information responsive to the attack to a third party.

Abstract: Data detection techniques involve: determining count features respectively corresponding to a plurality of time points based on data blocks of a storage object; determining, for a first time point among the plurality of time points and according to the count features respectively corresponding to the plurality of time points, a plurality of corresponding correlation coefficients between a count feature for the first time point and count features for the other time points among the plurality of time points; and determining, if a score calculated according to the plurality of correlation coefficients is less than a predetermined threshold, that the storage object corresponding to the first time point is under attack. Accordingly, a storage system can be protected, and the point in time when it is attacked by malware can be quickly located, which improves checking efficiency, saves checking costs, and helps users to quickly recover stored data.

Abstract: Techniques for data migration involve obtaining, at an edge device, a bitmap in a local memory, the bitmap including a plurality of parameter values. The techniques further involve updating, in response to a data block being written or read, a data heat value of the data block based on the bitmap. The techniques further involve migrating the data block between the edge device and a cloud device based on the updated data heat value. Accordingly, there is a solution for migrating data blocks between an edge device and a cloud device, such that data with a high data heat value is stored at the edge device and data with a low data heat value is stored at the cloud device, thereby improving the performance of a storage system.

Abstract: Techniques for snapshot classification involve generating a plurality of snapshots of a storage system at multiple times; clustering the plurality of snapshots into a first group of snapshots and a second group of snapshots based on data features of the plurality of snapshots; and determining a clean snapshot among the plurality of snapshots based on a comparison between the generation time of the first group of snapshots and the generation time of the second group of snapshots. Accordingly, a clean snapshot and a damaged snapshot can be automatically and quickly distinguished, thereby improving the efficiency of finding a clean snapshot and helping a user recover data quickly.

Abstract: Techniques for data processing involve: performing, according to a determination that a programmable circuit receives user-written data, data feature detection by the programmable circuit on a first part of the user-written data. Such techniques further involve: replacing, in response to the first part matching a predetermined data feature, the first part by the programmable circuit with a data representation corresponding to the predetermined data feature. In addition, such techniques involve: writing the data representation to a cache region of a storage system. Accordingly, such techniques can save cache resources of a storage system and can save CPU of the storage system, thus avoiding the latency due to data feature detection and improving the user experience.

Abstract: The present disclosure relates to an alloy wire rod and a preparation method and application thereof. The alloy wire rod is made of a tungsten alloy, and the tungsten alloy contains tungsten and an oxide of cerium. The alloy wire rod has a wire diameter of equal to or less than 100 ?m; and the alloy wire rod has a tensile strength of greater than 3,800 MPa. The wire diameter of the alloy wire rod is equal to or less than 60 ?m; the diameter of a push-pull core wire of the alloy wire rod is less than 350 ?m; the elastic ultimate strength of the alloy wire rod is greater than 2,500 MPa; and the tensile strength of the alloy wire is greater than 4,200 MPa. In the present disclosure, the alloy wire rod having ultra-high strength and good toughness is obtained by doping an oxide of cerium.

Abstract: A method of performing synchronous replication from a primary storage system apparatus (PSSA) to a secondary storage system apparatus (SSSA) is provided. The method includes (a) in response to write requests received by the PSSA, (i) calculating metadata changes by the PSSA for accommodating the write requests, (ii) generating, by the PSSA, metadata journal log entries that describe the metadata changes, and (iii) mirroring the metadata journal log entries from the PSSA to the SSSA; (b) regenerating the metadata changes by the SSSA based on the metadata journal log entries mirrored from the PSSA to the SSSA; and (c) writing the regenerated metadata changes to persistent storage of the SSSA. A method performed by the SSSA is also provided. An apparatus, system, and computer program product for performing similar methods are also provided.

Abstract: Techniques perform file storage. Such techniques involve determining a candidate allocation unit (AU) size for storing data based on a statistic of an AU size matching with a write request during a statistic period, and storing a new file using the determined candidate AU size to store the new file. Accordingly, storage space is saved and reading and writing performance are improved.