If i have more than 100k rows in a table and i want to display it in the Record. As it is a huge data how can i handle it to maintain the performance of the record.
Hi Vikas, Suppose i have already applied the filter and i have more than 100k of rows than how will i handle it? Thanks
Depending on the expression applied, using the default filters can actually perform badly as the expression is applied per row. Thus, using a grid for the record list and paginating to 25 rows won't necessarily help.
Do you expect the table to grow from 100k records? If so (and maybe regardless of this) I'd suggest using a service (expression) backed record as documented here; they are a fair bit more complex to implement but you gain greater control over the search functionality and over what is actually retrieved by using a query entity expression. You can also use a (well-written) database view to join various tables together and produce a view that matches the record list you want to display - just make sure you include the things you wish to filter on as well!
I have seen service backed records handle anything from 500k to 1m rows when used correctly with a view and a properly indexed set of database tables.
What is the performance of the stored procedure like? Are you stuck with a stored procedure or could you use a view or materialised view? Once you've tuned the stored procedure / view, make sure you're returning only the fields required for the record list view as well as a correctly paged datasubset (rather than returning everything and then paging it within the source expression). As I understand it, it is true to say that the entity backed record fetches only the data you required but it's how it does that fetch that can be a problem. For example, if you're returning 25 rows of a 1 million row dataset, the default filter can be written in such a way that it has to be applied per row rather than as a simple where clause. Obviously with an entity backed record, we don't have control over how this happens - but with a service-backed record we do. In fact, when using a service-backed record its often a good idea to move the default filter to inside the source expression so that you have control over how the filter is applied and can make sure it is done efficiently.
Discussion posts and replies are publicly visible
© 2020 Appian. All rights reserved.