Luis G Cabrera

US Patent:

6343341, Jan 29, 2002

Filed:

Aug 20, 1999

Appl. No.:

09/378054

Inventors:

Luis Felipe Cabrera - Bellevue WA
ATM Shafiqul Khalid - Bellevue WA
Stefan Robert Steiner - Issaquah WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06F 1202

US Classification:

711111, 711112, 711114, 711154

Abstract:

A system, a method and a program product provide access to variable-length data segments on a sequential access storage medium using an iterative search for a specified data offset based on estimates of media offsets. The data is recorded in variable-length data segments aligned to predetermined alignment intervals. Each data segment has a predetermined signature field to identify the header of each data segment. Session data recorded on the medium so as to prevent session data that matches the signature field from being aligned with the predetermined alignment interval. Approximate or default parameters for data segment size are used for each estimate and are updated with each iteration based on actual data derived from data segment headers on the medium.

US Patent:

6349313, Feb 19, 2002

Filed:

Aug 31, 2000

Appl. No.:

09/652347

Inventors:

Oshoma Momoh - Seattle WA
Gary D. Kimura - Kirkland WA
Luis Felipe Cabrera - Bellevue WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06F 1730

US Classification:

707205, 707100

Abstract:

A processing system processes an I/O request and, upon interruption of the I/O request, identifies an owner of an active mount point or another entity, and then resumes processing the I/O request under the control of the identified owner. For instance, when an I/O request processed by an I/O system encounters an active mount point, the processing is interrupted as control is turned over to an active mount point driver. The active mount point driver assumes responsibility for processing the I/O request. In processing the I/O request, the active mount point driver decides what action should be taken to further processing of the I/O request. To further processing of the I/O request, the active mount point driver may utilize information from a variety of sources as well as make requests to other processes, drivers, systems, subsystems, devices, and so forth. Such actions may include those normally associated with an I/O request such as retrieving and mounting appropriate media as well as actions not normally associated with an I/O request. This technique allows arbitrary actions to be performed when grafting the name space of one device into the name space of the same or another device using active mount points.

US Patent:

6421684, Jul 16, 2002

Filed:

Aug 31, 2000

Appl. No.:

09/652525

Inventors:

Luis Felipe Cabrera - Bellevue WA
Norbert P. Kusters - Woodinville WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06F 1730

US Classification:

707200, 707202, 707205

Abstract:

Information regarding volume mount points hosted by a logical volume are stored on the physical device underlying the logical volume so that the relationships between the host logical volume and target logical volumes mounted on the volume mount points can be reconstituted when the system containing the logical volumes is rebooted, when the underlying physical devices are moved with the system, and when the logical volumes are transported to a different system. A data structure stored on the physical device contains the directory name of the volume mount point and the unique identifier and a globally unique mount name of the target logical volume mounted at the volume mount point. When the target logical volume is present in the system, symbolic links are created to relate the volume mount point name to a device name for the target logical volume so that pathnames containing the directory junction name are resolved correctly. If the target volume is not present in the system, the corresponding symbolic link does not exist, so an incorrect logical volume cannot be mounted onto a volume mount point. Furthermore, because the logical volumes contain the directory junction information, the namespace representing the logical volumes is selfdescribing so that neither user knowledge nor intervention is required to reconstitute the namespace.

US Patent:

6453325, Sep 17, 2002

Filed:

Feb 3, 1997

Appl. No.:

08/794691

Inventors:

Luis Felipe Cabrera - Medina WA
Chandrasekaran Mohan - San Jose CA
Inderpal Singh Narang - Saratoga CA

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 1200

US Classification:

707204, 707200, 707201, 707203

Abstract:

In a database system with linkage between data in the database system and files in a system for filing data which is external to the database system (âthe filing systemâ), backup and restoration of the database are coordinated with the filing system. Backup of a referenced file is initiated when the file is linked to the database system. The file backup is asynchronous to the linking process. When database backup occurs, all unfinished file backups are accelerated and completed before the database backup is declared successful. When a database is restored to a state with reference to files in a file manager, the database system causes the file manager to ensure that referenced files are linked to the database system.

US Patent:

6490666, Dec 3, 2002

Filed:

Aug 20, 1999

Appl. No.:

09/378050

Inventors:

Luis Felipe Cabrera - Bellevue WA
Ravisankar V. Pudipeddi - Redmond WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06F 1206

US Classification:

711161, 711154, 710 56, 707204, 707103

Abstract:

A system, a method, and program products for buffering data from a file in a hierarchical data storage system allocates data buffers and buffer management structures in memory to optimize performance of no recall requests. Buffer management structures, such as buffer headers and hash queue headers, are used to optimize performance of insert, search, and data buffer reuse operations. Buffer headers are managed in a least-recently-used queue in accordance with a relative availability status. Buffer headers are also organized in hash queue structures in accordance with file-based identifiers to facilitate searching for requested data in the buffers. Data buffers can be used to buffer different data blocks within the same file and can be recycled to buffer data from other data blocks and other files from the secondary storage device. Data in a data block may be reread by the requesting process or by other processes as long as the requested data remains valid. Lock fields are used to coordinate multi-thread and multi-user accesses.

US Patent:

6496839, Dec 17, 2002

Filed:

Jun 12, 1998

Appl. No.:

09/096540

Inventors:

Luis Felipe Cabrera - Bellevue WA
Norbert P. Kusters - Woodinville WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

C06F 1200

US Classification:

707203

Abstract:

Persistent names for logical volumes in a computer system are used to enable symbolic links between the persistent names and non-persistent device names that identify the logical volumes to the lower layers of the operating system during a single boot session. A persistent unique mount name for each logical volume is associated with a unique volume identifier for the logical volume in a persistent mount manager data structure. Optionally a persistent drive letter can be assigned to the volume. Higher layers of the operating system and user applications address a volume through the persistent mount name as well as any assigned drive letter. When the system is rebooted, the data structure entries identified by the unique volume identifiers of the arriving logical volumes are used to reconstruct the symbolic links so that references to the persistent mount name, and optional drive letter, will resolve to the correct non-persistent device name, thus guaranteeing consistency in addressing the logical volumes across boot sessions. When the system undergoes physical reconfiguration, an existing persistent mount name is associated with a different non-persistent device name if the unique volume identifier is present in the data structure. In this fashion, logical volumes can be removed and restored in the computer without the knowledge of higher layers of the operating system and user applications.

US Patent:

6535998, Mar 18, 2003

Filed:

Jul 26, 1999

Appl. No.:

09/361481

Inventors:

Luis Felipe Cabrera - Bellevue WA
Kartik N. Raghavan - Seattle WA
Glenn A. Thompson - Redmond WA
Wesley Witt - Redmond WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

H02H 305

US Classification:

714 15, 714 2

Abstract:

A method and system for recovering from a system failure wherein the failed system is restored to a new system that has different hardware. Hardware state is preserved during a backup process, and following a failure, a restore is performed to the extent possible using that hardware state but on a system having a different hardware configuration. Rules are provided for handling the differences through selective merging, arranging, and replacement of data, with the logic and work performed transparently to the user. Hardware state includes hard disk configuration information, the location (partition) of the operating system, devices installed on the system and any additional drivers to load. When restored, for hardware that is identical, the hardware state is restored as specified in the file. If the hardware state is not identical, then a set of rules are used to restore the hardware state.

US Patent:

6553387, Apr 22, 2003

Filed:

Nov 29, 1999

Appl. No.:

09/450300

Inventors:

Luis Felipe Cabrera - Bellevue WA
Catharine van Ingen - Berkeley CA
Norbert P. Kusters - Woodinville WA
Rod N. Gamache - Issaquah WA

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06F 1200

US Classification:

707200, 707 1, 707204, 707100, 707103, 711114, 711167, 711170

Abstract:

Volume configuration data management systems and methods are described. A logical volume provider maintains an epoch number on each of the extents comprising a volume. Upon each change in the configuration of the volume, the epoch number on each of the currently online extents is updated. When a volume is discovered and a request is made to bring the volume online, the logical volume provider can compare the epoch numbers on the extents to determine if the volume data is consistent across each of the extents, or if an extent contains stale data due to a transitory outage.