By using rollup we can generate aggregate values like sum ,count…..etc where as by using scan we can generate cumulative summaries of records. In rollup the finalize function is applied to a group of records where as in scan it is applied to every record.

Rollup is used to aggregate data .

Scan is for successive total and is used to  produce summary(cumulative) of data.

guys,

Rollup:

 1)generally rollup is used for aggregation purposes such as sum,avg,count,min,max etc.

2)It can process grouped or ungrouped records.while processing ungrouped records it maximises the performance by keeping the intermediate results in main memory

3)It has more control over record selection,grouping and aggregation than aggregate component.

4)it is a multistage component

Scan:

1)scan is used to produce cumulative summary of records such as year-to-date totals for groups of input records

2)it is a multistage component

A multistage component is a componet which transforms the input records in 5 levels,such as input selection,temporary initialization,processing ,output selection,and finalize.for each level it is treated as a transform and a dml is written for each stage.

  hey ,not but the least we can have "SCAN WITH ROLLUP" component to cater the needs of both scan and rollup.

yuppy,

Mukund