Symantec Corporation

Veritas Technologies LLC

Veritas US IP Holdings LLC

Thomas Hartnett

Xianbo Zhang

Weibao Wu

Satyajit Gorhe Parlikar

Adonijah Park

Michael John Zeis

Benjamin Potvien

The present disclosure provides for defragmenting deduplicated data, such as one or more backup image files, stored in a deduplicated data store. A defragmentation module can be implemented on a deduplication server to reduce fragmentation of backup images and improve processing time for restoring a backup image. A defragmentation module can be configured to defragment a backup image file by migrating portions of data of the backup image file that are stored in various containers at non-contiguous locations throughout deduplicated data store. A defragmentation module can contiguously write the portions to one or more containers, which are stored at one or more new locations in the deduplicated data store. A defragmentation module can be configured to evaluate whether portions of a backup image file meet criteria for defragmentation. A defragmentation module can also be configured to update location information about the portions that are migrated to the new container(s).

Constrained backup image defragmentation optimization within deduplication system

A system and method for caching fingerprints in a client cache is provided. A data object that comprises a set of data segments and describes a backup process is identified. Thereafter, a request referencing the data object is made to a deduplication server to request that a task identifier be added to the data object. If the deduplication server is able to successfully add the task identifier to the data object, then an active identifier is added to each data segment from the set of data segments in a cache that is within a client system.

Processes and methods for client-side fingerprint caching to improve deduplication system backup performance

Systems and methods for providing efficient storage and retrieval of data are disclosed. A two-level segment labeling mechanism may be employed to ensure that unique data segments from particular backup data sets are stored together in a storage container. The two-level segment labeling may facilitate preservation of the relative positions of segments within the backup stream during compaction operations. Also, backup data restoration performance may be improved by use of multiple read threads that are localized to particular storage containers.

Efficient data storage and retrieval for backup systems

Method and apparatus for allocating resources among backup tasks in a data backup system is described. One aspect of the invention relates to managing backup tasks in a computer network. An estimated resource utilization is established for each of the backup tasks based on a set of backup statistics. A resource reservation is allocated for each of the backup tasks based on the estimated resource utilization thereof. The resource reservation of each of the backup tasks is dynamically changed during performance thereof.

Growing community of inventors

Saint Paul, MN, United States of America

Thomas Hartnett