ADIC FileServ Version 4.0 600716 Rev A manual Clustering

AdministrativeTasks

	File spanning is enabled or disabled for each DataClass group
	by the fsaddclass command. The fsmodclass allows
	modification of DataClass group processing parameters while
	the fsfileinfo command reports all the locations of the
	spanned file.
Clustering	FileServ supports file clustering. File clustering is the
	accumulation of files associated with a DataClass group. When
	these files are stored on media, they are seen as one logical file.
	Any file in a cluster is retrieved as any other non-clustered
	FileServ file.
	File clustering can be defined when adding a DataClass group
	to the file system (fsaddclass) or by the fsmodclass
	command. File clustering offers two advantages over individual
	file storage. The first advantage is the media space saved when
	clustering. The amount of space consumed by a cluster of files
	is less than the space used if files are stored individually. The
	second advantage is that the amount of time to store a cluster of
	files is less than the sum of the time required to store each file
	separately in the cluster. Clustered file information is sent in
	one continuous stream of data to the drive for storage on media.
	Unclustered files require the drive to stop, insert tape marks,
	file labels, and perform database updates for each file, which
	results in increased storage time.
	The system administrator controls the maximum size of the
	files to be clustered and the clustered file size for all supported
	media types.

2-24 Operating Tasks 600716 Rev A