Emc Corporation

Emc IP Holding Company LLC

Michael Paul Wagner

Alan Lee Taylor

David Haase

Paul Thomas McGrath

Michael C Brundage

Michael Lee Burriss

William Paul Hotle

Dave W Harvey

Sriram Nagasubramanian

David W Harvey

Dennis Thomas Duprey

Peter John McCann

Steven A Morley

Somnath A Gulve

Charles Christopher Bailey

Robert F Goudreau

Jackson Brandon Myers

Miles Aram DeForest

Chung-Huy Chen

J Brandon Myers

Dean A Herington

Michael J Noeth

Bruce Allen Keesee

Saurbh M Pathak

Techniques are presented for implementing a scalable approach to keeping track of the metadata validity of persistently-stored metadata for storage objects. In contrast with prior approaches, improved techniques only store a list of the invalid (or quarantined) objects rather than all objects regardless of validity of their metadata. Under normal circumstances, only a small number of storage objects are invalid at any given time, as the system generally takes steps to repair the invalid objects in the normal course of operation. Thus, the number of invalid objects tends to be small, barring some catastrophic system failure. Only a small amount of expensive privileged persistent storage is therefore needed. A health level may also be used to indicate many failures, indicative of catastrophic system failure.

Quarantining storage objects using a quarantine list

A technique for managing copy sessions in a data storage system supports active and inactive copy-session states and reclaims memory for inactive copy sessions to reduce their memory footprints. When a copy session transitions from the active state to the inactive state, the copy session persists certain copy-progress data in non-volatile storage and frees structures in memory previously storing that copy-progress data. The inactive copy session continues to run but performs no copying and consumes less memory. When the copy session later transitions back to the active state, the copy session reads the copy-progress data back into memory and resumes copying.

Managing copy sessions in a data storage system to control resource consumption

Techniques for processing requests are described. A first thread is selected for execution. Only a single thread is allowed to execute at a time. Each thread is associated with a queue of requests to be processed by the thread. A first request is selected from the queue of first thread that performs first processing to service the first request. A service time classification for the first request is determined in accordance with criteria that includes a runtime determination of what resource(s) are used in servicing the first request. It is determined, in accordance with the service time classification, whether to allow the first thread to continue execution and process a second request from the queue of the first thread. If the first thread is allowed to continue execution, second processing is performed by the first thread to service the second request. Otherwise, a second thread is selected for execution.

Using request service time as fairness heuristic in locking techniques

A method, computer program product, and computing system for receiving, on a first storage processor, a configuration IO request concerning a data array coupled to the first storage processor. The configuration IO request is provided to the data array for execution. A failure indication that the configuration IO request failed to execute is received. In response to receiving the failure indication, a status indicator is determined for a second storage processor coupled to the data array. Whether to reset the first storage processor is determined based, at least in part, upon the status indicator of the second storage processor.

System and method for storage management

A method is used in managing database access. A framework is provided for translating versioned data between memory and persistent storage. Based on the framework, read and write requests that access records in a database are translated. Based on the framework, it is made to appear that the database is a different version of the database.

Managing database access

Methods, systems, and computer program products for preserving atomic writes to disk arrays are disclosed. According to one method, portions of a file are stored in paired locations on a disk array such that a first location of a location pair for a file portion stores scratch data for the file portion and a second portion of the location pair for the file portion stores valid data for the file portion. A bitmap is stored on the disk array. The bitmap has bits indicating whether valid data for each file portion resides in the first or the second location for its respective location pair. A write request to write data to a portion of the file being less than the entire file is received. The bitmap indicates that valid data for the portion of the file to be written to resides in a first memory location of one of the location pairs on the disk array. In response to the write request, the bitmap is copied to memory, the data for the portion of the file requested to be modified is copied from the first location of the one location pair to a second location of the one location pair on the disk array without copying the entire file and modifying the data in the second location based on the data to be written. The bitmap stored in memory is modified to indicate that valid data for the portion of the file resides in the second location of the one location pair and that scratch data for the portion of the file resides in the first location on the disk array. A request to save the modification is received, and, in response to the request to save the modified data, the bitmap stored on the disk array is replaced with the modified bitmap.

Methods, systems, and computer program products for preserving atomic writes using leaves and bitmaps

An architecture and apparatus for atomic operations across multi-volume or multi-LUNs in a data storage environment. It is difficult to manage data storage and replication operations occurring across such multi-LUNs. The architecture and apparatus embodiments of the invention overcome these deficiencies and facilitate data storage management and replication operations through library initializations in response to host I/O requests to complete atomic operations. Accordingly, the invention ensures the integrity of the data being managed without introducing significant cost or overhead.

Architecture and apparatus for atomic multi-volume operations

A method system and program product for atomic operations across multi-volume or multi-LUNs in a data storage environment. It is difficult to manage data storage and replication operations occurring across such multi-LUNs. The method system and program product embodiments overcome these deficiencies and facilitate data storage management and replication operations through library initializations in response to host I/O requests. Accordingly, the invention ensures the integrity of the data being managed without introducing significant cost or overhead.

System and method for atomic multi-volume operations

A method allows point-in-time viewing of session data for multiple sessions being stored on a disk. Session bits are set for the session data in memory which points to the data at any particular point in time, and this can be done for multiple sessions. In order to make the session data persistent, separate session bits are stored in a portion of a disk reserved for such session bits. Thus, in the event of failure where session bits which are stored in volatile memory such as RAM, are lost, the duplicate copy set in the non-volatile memory can be used to recreate the session once the failure is corrected.

Method of allowing point-in-time view of data on a disk

A method of creating point-in-time views of data on a disk involves initiating from a host at least one session of writing data to a disk. For each session, an entry is created and stored on an array on the disk. The sessions are assigned unique IDs to each, and provide indications where data for a particular session may be found on the disk.

Method of allowing point-in-time view of data on a disk using a map on cache disk

A system, device, and method for providing mutual exclusion for computer system resources utilizes a lock mechanism in which a process that does not own a particular resource creates a lock for the resource upon receiving an inquiry from another process. The lock may be a full instantiation of a lock or a partial instantiation of a lock. In either case, the lock indicates, among other things, the owner process for the resource. When the non-owner process requires access to the resource, the non-owner process uses the lock to determine the owner process and sends a request to the owner process requesting access to the resource.

Growing community of inventors

Raleigh, NC, United States of America

Michael Paul Wagner