Hp Storage Mirroring Software User Manual download pdf (Page 12)

Storage Mirroring Replicate overview Page 11 of 214

Atom file-level deduplication

The Atom deduplication feature of Storage Mirroring Replicate provides the ability to

deduplicate protected data on the repository server. Deduplication is the process of

finding duplicate files within a data set and storing those files only once to reduce

storage consumption.

Atom finds duplicate data at the file level by inspecting file contents, generating an index

entry (technically known as a fingerprint), then comparing that fingerprint to a database

of fingerprints for the entire data set. When exact file duplicates are found, the contents of

the file are moved to a storage bin and all instances of the file are modified to reference

the single file in the storage bin through the use of reparse points. Because only file

contents are compared and deduplicated, files do not have to have matching file names

or attributes to be identified as duplicates . Each instance of a deduplicated file will

retain unique file attribute and filename information.

Deduplication occurs in real-time after the data is written to the Storage Mirroring

Replicate repository server. The Atom File-Level Deduplication Service initially scans

through the data set looking for duplicate files, then it monitors the NTFS change journal,

in real-time, looking for new or modified files as candidates for deduplication.

1 2 ... 7 8 9 10 11 12 13 14 15 16 17 ... 214 215

No comments

Hp Storage Mirroring Software User Manual Page 12