Well, this statement sounds like marketing-speak, and, by itself, it's not correct. Whether a database provides a relational or XML-based access is orthogonal to declustering strategies used for its storage management. SQL databases have been using fairly sophisticated data declustering techniques for many years. In addition, SQL databases have also been available in parallel versions, where declustering occurs on a set of processors (or nodes), not just across disks in a single node. In addition, seek time is only one factor in retrieving data -- for instance, having many disks bound to one bus (e.g., a SCSI bus) could cause that bus to become the bottleneck, etc., So, that statement is really an oversimplification.
Frank On Mon, 13 May 2002, White Shadow wrote: > One of the vendors of Native XML database has written the following > advantage of Native XML over the Realtional databases. > > "The major bottleneck in retrieving data from database is the seek time. XML > data can be clustered into one sector on a disk drive to greatly reduce the > seek time when searching for individual elements, thus speeding performance > times. In a relational database, each of the data elements to be searched > will be in different sectors of the disk, thus taking a certain time to > access each element, thus increasing time spent seeking the data." > > Is this argument Correct? Can someone elaborate how can XML data be > clustered into one sector and if it is possible,then why cannot be the > relational data clustered into one sector? > > Thanks > > Arshad > > _________________________________________________________________ > Join the world�s largest e-mail service with MSN Hotmail. > http://www.hotmail.com >
