公开(公告)号：DE112013006723B4

公开(公告)日：2016-10-13

申请号：DE112013006723

申请日：2013-12-16

Applicant: IBM

Inventor： HAUSTEIN NILS ,  SEIPP HARALD ,  TROPPENS ULF ,  WINARSKI DANIEL J

IPC: H04L9/16 ,  G06F12/00 ,  G06F13/00 ,  G06F17/30 ,  G09C1/00 ,  H04L47/31

Abstract: Verfahren zum Verwalten eines Datenpakets, wobei das Verfahren die Schritte aufweist: dass ein erster Computer ermittelt, dass eine Auslastung eines Netzwerks größer als ein Netzwerkauslastungs-Schwellenwert ist, wobei das Netzwerk einen ersten Computer und einen zweiten Computer enthält; auf Grundlage dessen, dass die Auslastung des Netzwerks größer als der Netzwerkauslastungs-Schwellenwert ist, dass der erste Computer in einen Deduplizierungsmodus wechselt und anschließend Schritte in dem Deduplizierungsmodus ausführt, die enthalten: dass der erste Computer eine Benachrichtigung an den zweiten Computer sendet, dass der erste Computer in den Deduplizierungsmodus gewechselt hat; dass der erste Computer ermittelt, ob eine Auslastung einer Zentraleinheit (CPU), die in dem ersten Computer enthalten ist, größer als ein CPU-Auslastungsschwellenwert ist; dass der erste Computer eine erste Hashfunktion statt einer zweiten Hashfunktion auswählt, wenn die Auslastung der CPU größer als der CPU-Auslastungsschwellenwert ist, und die zweite Hashfunktion statt der ersten Hashfunktion auswählt, wenn die Auslastung der CPU kleiner als der oder gleich dem CPU-Auslastungsschwellenwert ist; dass der erste Computer einen Hashauszug des Datenpakets durch Anwenden der ausgewählten ersten oder zweiten Hashfunktion ermittelt; dass der erste Computer ermittelt, ob sich der Hausauszug in einer Sender-Hashtabelle befindet, die mit dem ersten Computer verbunden ist; wenn sich der Hashauszug in der Sender-Hashtabelle befindet, ohne dass der erste Computer das Datenpaket zu dem zweiten Computer sendet, dass der erste Computer den Hashauszug, einen Index, der auf den ...

公开(公告)号：GB2525346A

公开(公告)日：2015-10-21

申请号：GB201513161

申请日：2013-12-16

Applicant: IBM

Inventor： HAUSTEIN NILS ,  SEIPP HARALD ,  TROPPENS ULF ,  WINARSKI DANIEL J

IPC: H04L9/32 ,  H04L47/31

Abstract: An approach for managing a data package is provided. Network utilization is determined to exceed a threshold.A sender computer determines a hash digest of the data package by using a hash function selected based on central processing unit utilization.If the hash digest is in a sender hash table, then without sending the data package, the sender computer sends the hash digest and an index referring to the hash digest so that a recipient computer can use the index to locate a matching hash digest and the data package in a recipient hash table.If the hash digest is not in the sender hash table, then the sender computer adds the data package and the hash digest to the sender hash table and sends the data package and the hash digest to the second computer to check the integrity of the data package based on the hash digest.

公开(公告)号：GB2507881A

公开(公告)日：2014-05-14

申请号：GB201318696

申请日：2013-10-23

Applicant: IBM

Inventor： AKELBEIN JENS-PETER ,  HAUSTEIN NILS

IPC: G06F17/30 ,  G06F19/28

Abstract: Data files are stored in a file system (10), wherein the file system (10) provides a plurality of reference data files (12), wherein each reference data file (12) represents a group of similar data files (14). The method comprises the steps of (i) creating a new data file (14); (ii) associating the new data file (14) with one reference data file (12) thus defining an associated reference data file (12) of the plurality of reference data files (12); (iii) informing the file system (10) about the association of the new data file (14) with the associated reference data file (12); (iv) compressing the new data file (11) using the associated reference data file (12); (v) storing the compressed data file (16) together with the information about the association with the associated reference data file (12). The invention can be applied to storing genomics data or seismic exploration data.

公开(公告)号：AT510280T

公开(公告)日：2011-06-15

申请号：AT08774530

申请日：2008-06-30

Applicant: IBM

Inventor： BATES ALLEN ,  HAUSTEIN NILS ,  KLEIN CRAIG ,  WINARSKI DANIEL

IPC: G11B7/0065

Abstract: A holographic storage drive and control of a holographic storage system are configured to write at least a group of holograms in a predetermined pattern in the holographic storage medium, and to write a directory hologram which relates to the holograms of the group to form an anchor location of the predetermined pattern. Further, a directory comprising the directory hologram stored in a memory, and the control is configured to initiate a read operation of at least one hologram of a group with an access for the directory hologram of the group at the anchor location; to read the accessed hologram, employing a matched filter to cross-correlate the read accessed hologram with an ideal version of the directory hologram derived from the directory stored in the memory; and to determine whether the read accessed hoIogram is the directory hoIogram of the group.

公开(公告)号：DE602008006092D1

公开(公告)日：2011-05-19

申请号：DE602008006092

申请日：2008-07-28

Applicant: IBM

Inventor： BATES ALLEN KEITH ,  HAUSTEIN NILS ,  KLEIN CRAIG ANTHONY ,  WINARSKI DANIEL JAMES

IPC: H03M13/23

公开(公告)号：DE602008001891D1

公开(公告)日：2010-09-02

申请号：DE602008001891

申请日：2008-04-11

Applicant: IBM

Inventor： WINARSKI DANIEL JAMES ,  BATES ALLEN KEITH ,  KLEIN CRAIG ANTHONY ,  HAUSTEIN NILS

IPC: G11B7/0065 ,  G11B7/09 ,  G11B7/24 ,  G11B7/24038 ,  G11B7/24044

公开(公告)号：GB2531295A

公开(公告)日：2016-04-20

申请号：GB201418219

申请日：2014-10-15

Applicant: IBM

Inventor： HAUSTEIN NILS ,  STEINHAUER SABINE ,  PRAUSE THOMAS ,  UHL ANDREAS ,  MUELLER-WICKE DOMINIC

IPC: G06F11/20 ,  G06F11/14

Abstract: A data management system 10 comprises file system 110 controlling access by database application 119 to database container file 115 stored in first storage 121, backup client 110 connected to remote backup server 120, where a first version (801v1 in Figure 8) of the container file is saved, and a first inode comprising information on data blocks of the first version of the container file. The method comprises: creating a change tracking table 130-131 (803 in Figure 8) for the container file, and repeatedly adding an entry upon detecting an access to a data block, where the entry comprises an indication 146 of the data block and information 148 indicating the type of access, including at least one of read and write; pausing application execution; performing a snapshot of a storage volume containing the file system and container file in second storage 122; resuming the application; mounting the snapshot storage volume; recreating the file system in second storage; identifying entries 144 in the table associated with a write access; copying the associated data to the backup server; and updating the first inode, obtaining a second inode. A journaling function may be used for detecting an access. Incremental backup is facilitated.

公开(公告)号：GB2527296A

公开(公告)日：2015-12-23

申请号：GB201410671

申请日：2014-06-16

Applicant: IBM

Inventor： HAUSTEIN NILS ,  MUELLER-WICKE DOMINIC ,  MUELLER CHRISTIAN

IPC: G06F17/30 ,  G06F3/06

Abstract: Methods of versioning or backing up data a hierarchical storage management, HSM, system which includes at least a first storage tier and second storage tier managed by a HSM client and HSM server respectively, wherein the HSM server and the HSM client are interconnected via a network of the HSM system, wherein a data item stored in the HSM system is associated with a HSM status parameter that is stored in the HSM client, the HSM status parameter having a first, second or third value, the first value indicating a pre-migrated state of the data item in which the data item is stored in the first and second storage tiers, the second value indicating a migrated state of the data item in which the data item is stored in the second storage tier, the third value indicating a resident state of the data item in which the data item is stored in the first storage tier, wherein in case the HSM status parameter has at least one of the first and second values the data item is further associated with a predetermined initial object ID that is stored in the HSM client and the HSM server, the initial object ID corresponding to a first version of the data item, the HSM server being configured with a data structure comprising an object ID field comprising at least the initial object ID and a newly created parent object ID field comprising the initial object ID.

公开(公告)号：CA2624993C

公开(公告)日：2014-07-08

申请号：CA2624993

申请日：2006-10-11

Applicant: IBM

Inventor： HAUSTEIN NILS ,  KLEIN CRAIG ANTHONY ,  WINARSKI DANIEL JAMES

IPC: G11B7/24085 ,  G11B7/26

Abstract: A medium, system, and method are disclosed for a common data storage medium depression depth. An optical data storage medium comprises a plurality of spacer layers and data surfaces. A wavelength for a radiation beam such as the emission of a laser diode is identified. The index of refraction for a first spacer layer or substrate is also identified. The substrate is configured to transmit the radiation beam. The index of refraction of a second spacer layer that is configured to transmit the radiation beam is also identified. A depression depth for a plurality of pits for ROM media, or sector headers for recordable media, or grooves for recordable media on each data surface is substantially equal to the radiation beam wavelength divided by four times the average of the indexes of refraction of each spacer layer. In one embodiment, the average is an arithmetic mean of the indexes of refraction. In an alternate embodiment, the average is a harmonic mean of the indexes of refraction. The average may also be a geometric mean of indexes of refraction.

公开(公告)号：GB2507751A

公开(公告)日：2014-05-14

申请号：GB201220084

申请日：2012-11-07

Applicant: IBM

Inventor： AKELBEIN JENS-PETER ,  HAUSTEIN NILS

IPC: G06F17/30

Abstract: Data files are stored in a file system (10), wherein the file system (10) provides a plurality of reference data files (12), wherein each reference data file (12) represents a group of similar data files (14). The method comprises the steps of (i) creating a new data file (14); (ii) associating the new data file (14) with one reference data file (12) thus defining an associated reference data file (12) of the plurality of reference data files (12); (iii) informing the file system (10) about the association of the new data file (14) with the associated reference data file (12); (iv) compressing the new data file (11) using the associated reference data file (12); (v) storing the compressed data file (16) together with the information about The association with the associated reference data file (12).