International Business Machines Corporation

National Security Agency

David Allan Pease

Sandeep Madhav Uttamchandani

Arnon Amir

Mohit Saxena

Anurag Sharma

Rainer Richter

James Lee Hafner

Mario Blaum

Kenneth Fairclough Day, Iii

Maohua Lu

Mihail Corneliu Constantinescu

Elizabeth Suzanne Richards

Edward Gustav Chron

Abdullah Gharaibeh

Mingyuan Xia

Atsushi Abe

Kaladhar Voruganti

Douglas William Dewey

John Jay Wolfgang

Lucas Correia Villa Real

Mark Andrew Smith

Jayram S Thathachar

Karen Wolfe Brannon

Cheryl Ann Kieliszewski

Sangeeta T Doraiswamy

Joseph P McCloskey

Ryan John Minniear

Wenling Cai

Sridevi Koushik Kanagal Ramesh

Sudarshan M Srinivasan

Laura Challman Anderson

James Vernon Carlson

Wayne Isami Imaino

Kenneth Wayne Boyd

Ying Chen

Ronald Mraz

Pin Zhou

Philip Matthew Doatmas

Kenneth F Day, Iii

Toby Lyn Marek

David R Nowlen

Linda Marie Duyanovich

Dean Lee Hanson

Michael Leo Walker

Michael Anthony Richmond

Norman Iwao Hanami

Bernard John Lopez

Brian Biskeborn

Tohru Hasagawa

Paul Harold Hilton

Sundarshan Srinivasan

One embodiment provides a method including: collecting system log files comprising a plurality of log messages representing activity within the system; generating a plurality of message templates by (i) clustering the plurality of messages into groups having similar activity patterns and (ii) generating a message template for each of the groups; identifying message templates of interest and determining the message templates representing actual end-user events as opposed to system events, wherein the determining comprises (i) identifying a seed set of event types that represent known end-user events based upon identifying the event as being known to be initiated by a user and (ii) correlating the event to the known end-user event based upon a time correlation between the known end-user event and the event within the message template of interest.

Generation of end-user sessions from end-user events identified from computer system logs

Data deduplication for data storage tapes includes intercepting tape control commands for a single data storage tape. The intercepted tape control commands are modified for adding processing logic and parameters for placement of deduplicated file data on the single data storage tape. Deduplication metadata is written to a metadata portion of the single data storage tape. The deduplicated file data is written to a data portion of the single data storage tape based on the placement to increase read throughput for a deduplicated set of individual files and to reduce an average number of per-file gaps on the single data storage tape without re-duplicating deduplicated data for meeting optimization of individual file accesses.

Optimized data placement for individual file accesses on deduplication-enabled sequential storage systems

In one embodiment, a data storage system includes a tape drive configured to: write a plurality of files to a first partition of a magnetic recording tape; and write an index to a second partition of the magnetic recording tape, the index including information about locations of data of the plurality of files in the first partition of the magnetic recording tape. In another embodiment, a magnetic recording tape includes: a plurality of files to written to a first partition of the magnetic recording tape using a tape drive; and an index written to a second partition of the magnetic recording tape using the tape drive, the index including information about locations of data of the plurality of files in the first partition of the magnetic recording tape.

File index, metadata storage, and file system management for magnetic tape

One embodiment provides a method including: collecting, from a plurality of collaborative computing systems, system log files comprising a plurality of log messages representing activity within the system, wherein the activity within the system corresponds to system activity and end-user activity; generating a plurality of message templates, each of the plurality of message templates representing an event type by clustering the plurality of messages into groups having similar activity patterns; identifying, from the plurality of message templates, a subset of message templates representing potential end-user events; differentiating end-user events from system events, in the subset of message templates representing potential end-user events, comprising identifying a seed set of event types that represent known end-user events and correlating potential end-user events remaining after identifying the seed set of event types with the known end-user events; and grouping, based upon the differentiating, the event types representing end-user events into user sessions.

Embodiments relate to a system with multiple erasure codes, and selecting and encoding for a write file with one of the codes to mitigate costs associated with storage recovery. The codes include a fast recovery code for frequently accessed data and a higher storage efficiency code for less frequently accessed data. State data is tracked to ascertain frequency of access to the file. One of the erasure codes is dynamically selected based on the tracked data, with the focus of the code select to lower recovery costs, and the data is encoded with the selected erasure code. Accordingly, the original coding of the write file is subject to change based on the tracked state data.

Multiple erasure codes for distributed storage

Various embodiments relate to providing dynamic absolute-path symbolic links. According to one embodiment, indication of a symbolic link is received, which references a target, the indication including an absolute path of the target that includes a target mount point and a name of the target. A current mount point of the target is identified, and a verification is made that the target mount point is identical to the current mount point. When the target mount point is identical to the current mount point, a string length of the current mount point is determined, and an in-memory record of the symbolic link is created. The in-memory record includes the string length of the current mount point associated with the symbolic link.

Dynamic symbolic links for referencing in a file system

Data deduplication for data storage tapes includes intercepting tape control commands for a single data storage tape. The intercepted tape control commands are modified by adding processing logic and parameters for placement of deduplicated file data on the single data storage tape. Deduplication metadata is written to a metadata portion of the single data storage tape. The deduplicated file data is written to a data portion of the single data storage tape based on the placement to increase read throughput for a deduplicated set of individual files and to reduce an average number of per-file gaps on the single data storage tape without re-duplicating deduplicated data for meeting optimization of individual file accesses.

Embodiments relate to dynamically selecting an erasure code. State data is tracked to ascertain frequency of file access. One of at least two erasure codes are selected based on the tracked state data in order to lower data recovery cost. The erasure code may be selected as either a product code or a local reconstruction code. Each erasure code includes a mode that is either a fast code or a compact code. The fast code features a low recovery cost and the compact code features a low storage overhead for less frequently accessed data. Data is encoded with one of the selected erasure codes and one of the modes of the selected erasure code. Data blocks are dynamically converted between the fast and compact codes of the selected erasure code responsive to a workload change.

Embodiments of the invention relate to providing dynamic absolute-path symbolic links. According to one embodiment, indication of a symbolic link is received, which references a target, the indication including an absolute path of the target that includes a target mount point and a name of the target. A current mount point of the target is identified, and a verification is made that the target mount point is identical to the current mount point. When the target mount point is identical to the current mount point, a string length of the current mount point is determined, and an in-memory record of the symbolic link is created. The in-memory record includes the string length of the current mount point associated with the symbolic link.

In one embodiment, a method for writing data to a magnetic recording tape includes writing a plurality of files to a first partition of a magnetic recording tape using a tape drive, and writing an index to a second partition of the magnetic recording tape using the tape drive, the index including information about locations of data of the plurality of files in the first partition of the magnetic recording tape.

Embodiments of the invention relate to a para-virtual I/O system. A consistent para-virtual I.O system architecture is provided with a new virtual disk interface and a semantic journaling mechanism. The virtual disk interface is extended with two primitives for flushing and ordering I/O, both of the primitives being exported to para-virtual I/O drivers in a guest operating system. The ordering primitive guarantees ordering of preceeding writes, and the flushing primitive enforces order and durability. The guest drivers selectively uses both of these primitives based on semantics of the data being persisted from the para-virtual cache hierarchy to physical disk. The order of committed writes is enforced in order to enable a consistent start recovered after a crash.

Efficient and consistent para-virtual I/O system

A system according to one embodiment includes a tape cartridge, the tape cartridge comprising: a housing; a magnetic recording tape in the housing; and a non-tape nonvolatile memory coupled to the housing, the nonvolatile memory being for storing therein an index comprising file system information for a plurality of files stored on the magnetic recording tape. A method for storing data on a tape cartridge according to one embodiment includes writing a plurality of files to a magnetic recording tape of a tape cartridge; and writing an index to a non-tape nonvolatile memory of the tape cartridge, the index including information about locations of data of the plurality of files on the magnetic recording tape.

Primary storage media with associated secondary storage media for efficient data management

Data deduplication for data storage tapes comprises determining the read throughput of a deduplicated set of individual files on a single data storage tape, and determining a placement of deduplicated file data on a single data storage tape to reduce an average number of per-file gaps on the tape. Deduplicated file data is placed on the single data storage tape based on said placement to increase an average read throughput for a deduplicated set of individual files.

In one embodiment, a method includes loading a tape cartridge into at least one tape drive installed in an automated tape library, where a tape of the tape cartridge has at least two partitions; writing plurality of data blocks on a first of the partitions; and writing an index on a second of the partitions, wherein the index includes information about at least one of files and the blocks on the first partition.

Growing community of inventors

Redwood Estates, CA, United States of America

David Allan Pease