Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations IamaSherpa on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

"Distributed Database System" says___ But what does it mean?

Status
Not open for further replies.

someTimeOnly

Technical User
Oct 12, 2003
78
PK
hi
A book "Distributed Database System" says
“If a predicate influences how fragmentation is performed, (i.e., causes a fragment f to be further fragmented into, say, fi and fj ) then there should be at least one application that accesses fi and fj differently.”

What does
“then there should be at least one application that accesses fi and fj differently.”
mean?

Does it mean that
at least one application should be satisfied completely
by one of the new fragments.

1. Am I right?
“accesses fi and fj differently” is confusing me.

2. We can be required to make fragments once in a month in any organization. Sounds right. Doesn’t it?

 
Hi,

A distributed database is quite simply (removing all the academic waffle) a database system in which the data is split across multiple storage locations.
They could be in the same server but different RAID boxes, different servers in the same machine room or going right the other way - in a different continent.

The main reasons for fragmentation are:
* security (ie making sure only "local" users can fully access the data that they are entitled to, and not access data to which access should be denied to them).
* performance - again especially if there are geographically separate, accessing a server in the machine room downstairs is going to be quicker than one halfway across the planet.

Fragmentation can be done on a horizontal or vertical basis:

Horizontal fragmentation has certain fields kept on different sites, with the primary key replicated at each site.
For example, if there were fields A, B, C, D, E and F with A as the primary key,
A, B, C, D could be at site one and A, E, F at site two.
The primary key is of course used to match up the data across sites.

Vertical fragmentation has the entire structure at each site, but a subset of the data. For example, there may be 10000 records at site 1, but 50000 at site 2.
The method of working out which is better for you depends upon your circumstances.

John

 
hi

U have defined Horizontal fragmentation for Vertical fragmentation and Vertical fragmentation for Horizontal fragmentation.

Is Practically fragmentation automated or not( given some relations along whith some user applications' conditions to some algorithm and getting the desired fragments as an output)?
 
Hi,

Apologies for getting the definitions of horizontal and vertical fragmentation mixed up - maybe I should go back and read my distributed database uni textbooks. :)

As to whether it is automated or not, I think it depends very much on your system. My own "real world" distributed database systems experience is limited to tying an Oracle database as one fragment, and a text file as the rest (university assignment, 5 years ago) so I am not in a position to give an expert answer, but I'm sure somebody else here will know.

John
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top