International Business Machines Corporation

Michael Diederich

Erik Rueger

Thorsten Muehge

Rainer Wolafka

Ulf Troppens

Thomas Doerk

Frank Krick

Itzhack Goldberg

Lance Warren Russell

Neil Sondhi

Nikhil Khandelwal

Monty Christoph Poppe

Stefan Schmidt

Thomas Bueck

Frank Broede

Christoph Roth

According to one embodiment, a computer-implemented method for file handling in a hierarchical storage environment includes performing a file access notification process for determining files related to the first file based on enhanced metadata and a priority list defining a likelihood of possible access, in response to receiving a file access notification corresponding to access of a first file. The related files are placed in a highest level storage tier, and the priority list is updated.

Method for file handling in a hierarchical storage environment and corresponding hierarchical storage environment

Methods, computer program products, devices, and/or systems are provided that perform the following operations: receiving data from an incoming data stream; separating the data into data chunks and storing the data chunks in accordance with a selection configuration matrix in a set of different matrix buffers until the matrix buffers are filled up; processing the data chunks by multiple processing units, each processing unit associated to a matrix buffer of the matrix buffers via an I/O block; providing the processed data chunks as inputs to a neural network component; and classifying the data and filtering the data based on a classification of the data by the neural network component.

Classifying and filtering data from a data stream

Data may be handled based on compressibility (i.e., whether the data may be further compressed or is not further compressible). A supervised learning model may be trained using a set of known further compressible data and a set of known non-compressible data. Using these data sets, the model may generate weighting factors and bias for the particular data sets. The trained model may then be used to evaluate a set of unclassified data.

Decision-based data compression by means of deep learning technologies

A RAID (redundant array of independent storage devices) storage system where a file is stored at a physical protection level corresponding to a RAID protection level defined by metadata included in and/or with the file to be stored.

File level defined de-clustered redundant array of independent storage devices solution

Solutions are provided that use shared scan phases and scan output for various file-level incremental data analysis systems. In one embodiment, a shared scan phase is initiated for a plurality of files in a file system. During the shared scan phase, one or more rules are applied to the files in the file system to identify files on which to perform one or more operations. Shared scan output is created that includes information describing the identified files and operations to be performed on the identified files. Embodiments of the present invention can reduce the amount of time and computing resources that would otherwise be consumed by performing separate walkthroughs of a file system during separate scan phases.

Shared scan output in incremental data analysis systems

Provided is a system having a cluster of communication controllers, a method for modification of the latter, and a computer program product carrying computer executable code for execution of the method. Each communication controller is operable for providing network connections of the system with external computer systems using communication protocols of a first type and a second type. The first type is a failover tolerant communication protocol type. The second type is a failover non-tolerant communication protocol type. All network connections of each communication controller are disconnected during the modification of said communication controller. The method for modifying the cluster of communication controllers includes splitting the communication controllers into a first pool and a second pool.

Modification of a cluster of communication controllers

A method for file handling in a hierarchical storage environment includes initiating a new document notification process in response to receipt of a new document. The document notification process includes evaluating enhanced metadata of the new document via a relationship analyzing process to produce a priority list defining a likelihood of possible access, where the enhanced metadata includes a set of keywords and/or references from the new document. The method additionally includes storing the new document in a storage tier of a hierarchical storage environment according to the priority list.

A method and apparatus for increasing storage capacity on magnetic media. A dual-layer magnetic recording apparatus receives, a write operation instruction, bit position and bit value where the bit position includes a media surface position and magnetic media layer identifier. The dual-layer magnetic recording apparatus accesses a dual-layer magnetic media where the dual-layer magnetic media includes a first magnetic media layer with a thermally controllable coercivity, a second magnetic media layer and a separation layer. The magnetic layer identifier is determined to associated with be the first magnetic media layer or the second magnetic media layer. When the magnetic layer identifier is the second magnetic media layer, the bit value is written based on magnetically charging the second magnetic media layer and when the magnetic layer identifier is the first magnetic media layer the bit value is written based on heating and magnetically charging the first magnetic media layer.

Dual-layer magnetic recording

A mechanism is provided in a data processing system for securing data integrity in de-duplicated storage environments in combination with software defined native redundant array of independent disks (RAID). The mechanism receives a data portion to write to storage, divides the data portion into a plurality of chunks, and identifies a given chunk within the plurality of chunks for de-duplication. The mechanism increment a de-duplication counter for the given chunk and determines a RAID level for the given chunk based on a value of the de-duplication counter. The mechanism stores the given chunk based on the determined RAID level.

Protecting data integrity in de-duplicated storage environments in combination with software defined native raid

First, second, and third sets of addresses are created. The first set includes addresses registered in a name server; both the second and third sets include addresses not registered in the name server and that are disjoint. A first address of a first server that has failed and to which access is to be prohibited is moved from the first to the third set, is removed from the first server, assigned to a second server, and removed from the name server. Usage parameter values of the first address are monitored to determine whether at least one is below a value. If so, the first address is removed from the second server and moved from the third to the second set. Upon access to the first server no longer being prohibited, a second address of the second set is assigned to the first server and added to the name server.

Controlling access of clients to service in cluster environment

A method for processing a data object access request in a storage system having storage devices can include receiving an access request for accessing a first data object, the request being received by a controller that stores at least two consecutive data objects received by the storage system in different storage devices. The method may then include detecting that the first data object is corrupted, and determining, in response to the detecting, that a second data object has identical data as the first data object. Determining that the second data object has identical information as the first data object can include determining that a first metadata tag corresponding with the first data object and a second metadata tag corresponding with the second data object are identical. The second data object can then be provided in a response to the access request.

Using duplicated data to enhance data security in RAID environments

A mirrored storage system comprising a system controller coupled to a first storage system comprising a primary controller and a first storage and to a second storage system comprising a secondary controller and a second storage. The second storage contains an image of data stored in the first storage. The system controller receives data to be written to storage and sends the data to the primary controller of the first storage system and to the secondary controller of the second storage system. The system controller instructs the primary controller to write the data to the first storage of the first storage system in an ordered manner; and instructs the secondary controller to write the data sequentially to a data area on the second storage of the second storage system. Writing the data sequentially to the second storage improves the write performance.

Sequential write of random workload in mirrored performance pool environments

Performing maintenance to controllers in a network-attached storage system that includes two or more controllers, each controller including at one or more associated IP address. An online controller in the network-attached storage system to upgrade is identified. The IP addresses of the online controller to upgrade is moved to another controller in the network-attached storage system. Simultaneous with when the other controller is online, the online controller to upgrade is taken offline and upgraded the controller to upgrade.

Growing community of inventors

Mainz, Germany

Michael Diederich