About D/W and OLAP

Question

Hi All,
I am always confused and could't understand exact difference between Data warehouse & DB and also about OLAP & OLTP. Could some one help me on this plz.
Plz Send at least one URL which explains it with a good example.
Thanks,
Naresh
Edited by: 793302 on Mar 9, 2011 9:03 AM

Answer 1

hi ..
OLTP (online transaction processing) or operation system – Does transaction processing (responsible for daily operations of business), The data warehouse is specifically confined to query processing.
OLTP maintains current state (snapshot) of data. Does not maintain history of data changes. The data warehouse maintains history of data.
OLTP provides operator view enabling tactical operational decision making. The operator has to see day to day data, for example what is the daily inventory, how many tickets has been issued today, the software should be running up and perfectly so it should not stop the processes etc.
Data warehouse gives managerial view as the trend, behavioral analysis, performance with ad-hoc query feature. Say, it will see the daily inventory for last 30 days and find out that is the inventory adequate or less or more than the required.
OLTP data is organized by transactions aligning to the business processes, the data warehouse is by subjects and rolled up to a higher level than operational applications. For e.g. the daily inventory checking in operational system is not needed in data warehouse. Here it will see the last 30 days inventory levels or average of inventory levels in last 30 days. The data is rolled up. The roll up means seeing the data from one level up in the summary process.
OLTP has smaller database size as it does not maintain history; data warehouse has huge volume of data as it maintains history for longer time period. The smaller database of OLTP system means the older data say one year or two year old data is taken offline for e.g. kept in tape storage. The data is not deleted very fast. The very old data may be 5 or ten years old data can be deleted if the business permits. IN data warehouse the data for last 5-10 years is kept active in database. The old data is not kept active in OLTP system, since it will slow down the daily data processing operations.
OLTP has many operational users accessing the OLTP data at the same time (high user concurrency), the data warehouse has few users.
OLTP has volatile data with frequent insert, update and delete operations. Data warehouse data is static. Here the Incremental data load takes place and data is accessed in the form of querying / reporting. Data deletion is rare and only for specific requirement.
OLTP captures all the data elements required for business process transaction integrity and completeness, the data warehouse captures relevant data only from source systems. For e.g. in OLTP it is important to know who inserted the record or who deleted the record. In data warehouse it has almost no use.
In OLTP, system is not flexible to handle changes to the business process (like adding / removing new columns) as this impacts the front end application, middle layer business application logic and the back end database schema changes. data warehouse is flexible to the changing the business needs (adding new data elements, bringing new source data into DW).
mohan