Report Definition with 50K records
We have reference data with 50K+ rows and need to compare only one row with primary key match at one time. Currently we use report definition to fetch the rows and a datapage at node level to read the row. Is this an efficient way to do this? Would an activity with obj-browse be better? Is it ok to keep 50k+ records in memory? Also, how to use the parameter from data page to report definition to fetch only one row?
***Edited by Moderator Marissa to update platform capability tags****
Coforge DPA
GB
Hi
Two things here :
1. Your current approach should be fine.
2. You must be using this Data Page in some rule to fetch the details. Then how do you plan to generate the Primary key for that one record initially?
Or do you have a segregation based on Reference Data Type in your table?
You can surely achieve the parameterization with either of Options in Point 2. Confirm on that and then we can see what configuration to make.
Regards
Bhavya
Merkle
US
Hi Bhavya,
Thanks for replying.
So, to confirm, you think its not a issue with having such a large number of records in memory?
Because one thought is, this might have an impact on performance. Why do you think this should be ok? Please let me know.
Secondly, the reference table stores a list of all location details with locationcode as PK. We are getting this locationCode in the incoming json request that we can pass as parameter. Hope that helps you with your question.
Thanks!
-Hansa.
Coforge DPA
GB
Hi
When you say memory you mean "Clipboard" , if yes then thats not the preferred way to load 50k records at once.
Wherever you have to use Location details you can refer you data page as D_LocationDetails[LocationCode: <Value>] and in the report definition have location code as parameter and is "Equals" in comparison.
Regards
Bhavya
Pegasystems Inc.
IN
Hi,
Please refer below article,here they have clearly explained about it.
Regards,
Vinay Reddy
ANZ Banking Group Ltd
AU
Hi Hansa,
I would recommend using a data page of structure page because you are expecting only one output every time and source as lookup as you know the primary key of the record.
The scope can be a thread level as per your scenario.
With this approach, you will load each record as per need.