to improve the performance of aggregator we use sorted input option and use sorter t/r befor aggregator. But here we are increasing one more cache in our mapping i.e; sorter. So how can u convince that you are increasing the performance.?
- 3 Answers
- 7216 Views
- IBM, I also Faced
- E-Mail Answers
Answers were Sorted based on User's Feedback
Answer / koti
To do aggr. caliculations function it needs some time.. we
can reduce that time by providing sorted I/P. The time is
less which can takes to forward the rows from sorter to
aggregator rather than the time to do aggr. calculations
without sorted data.....
| Is This Answer Correct ? | 2 Yes | 0 No |
Answer / bhupal
if u wont use the sorted input the aggregator transformation will reads the records and keep until all the calculations are done.
if we provided the sorted input it reads the records and release the records reaching next first record of new group.
the integration service will creates the cache for only the sorter keys and only one cache in sorter t/r. so by providing the sorted input we get the performance advantage.
| Is This Answer Correct ? | 1 Yes | 0 No |
Answer / james
Sorted input in an aggregator transformation does not use
cache instead it uses the memory. Reading from memory is
faster than reading from cache.
| Is This Answer Correct ? | 0 Yes | 1 No |
Hi If i had source like unique & duplicate records like 1,1,2,3,3,4 then i want load unique records in one target like 2,4 and i want load duplicate records like 1,1,3,3 then can any body please send me what is th scnario. my mail i shek.inform@gmail.com
there is a product table prodid prodname price 100 cinthol 10 101 hamam 10 102 neem 20 103 cake 30 in the above table the price of some products are duplicated and some product prices are distinct we want to push the duplicated prices to one target and non-duplicated prices to other target without using expression and sequence generator transformation
Differnce between filter and router?
How many ways a relational source definition can be updated and what are they?
My questions is i create a two sessions for one mapping.but my requirement is if all number of source records are same as in target then execute first session or some rows are rejected due to t/r logic so session two was execute please clarify
What is the difference between OLTP and ODS?
How the dimensions will be loaded?
How will you update the row without using update statergy?
Hi, Can anyone just give me the list of all versions of Informatica ? Advance Thanks
Whatis the difference between View and Materialized View ?
i have a wf like wf-->s1-->s2-->s3-->s4 first start s1 later s2 later s3 here my session s3 have to run 3 times later start s4?
generate Unique sequence numbers for each partition in session with Unconnected Lookup ? Hi All, Please help me to resolve the below issue while Applying partitioning concept to my Session. This is a very simple mapping with Source, Lookup , router, and target. I need to Lookup on the target and compare with the source data, if any piece of data is new then Insert, and If any thing change in the existed data then Update. while Inserting the new records to the target table I'm generating sequence numbers with Unconnected lookup, by calling the maximum PK ID from the target table. The above flow is working fine from last one year. Now I wish to apply the Partitioning concept to the above floe(session) At source I used 4 pass through partitions.(For Each partition different filter conditions to pull the data from source) at Target I used 4 passthrough Partitions. it is working fine for some data, but for some rows for Insert Operation , it is throwing Unique key errors, because while Inserting the data it is generating the same sequence key twice. In detail : 1st row is coming from 1st partition and generated the sequence number 1 for that row. 2nd row is coming from 1st partition and generated the sequence number 2 for that row 3rd row is coming from the 2nd partition generated the sequence number 2 again for that row. (it must generate 3 for this row) the issue is becuase of generating the same sequence numbers twice for different partitions. Can any one Please help me to resolve this issue. While Applying partitions how can I generate a Unique Sequence numbers from Unconnected lookup for Each partitioned data. Regrads, N Kiran.
Teradata 
