1

The database is oracle 12c. The table contains 1 million rows.

I am running two sets of queries. The first one -

select * from table_name
select * from table_name where rownum <=10

Another:

select * from table_name fetch first 10 rows only

The first set does not run and is stuck for a long time while the second set runs just fine. I am not sure how to debug it.
I am sure a normal select does not fetch the whole table at the same time or does it? And when you cancel the query it takes a lot of time too (which I think is kind of okay).
There must be some setup issues but can’t figure that out. I need some help here. Thanks

EDIT: I had to flush Redis and then only things started working. I believe the long-running queries are getting stuck and then subsequent queries are sufferingl

2

It’s not a bug. Redash loads query results directly into your browser. If a query returns 1 million rows then I’d expect the entire browser tab to crash :stuck_out_tongue:

Why do you need to load so much information at once? Usually, result sets around 20mb are the sweet spot (though it varies from one browser to the next).

3

Ahh thanks for the reply. Can I put a default limit(say 1000 rows) on a plain Select * query - somewhere on the backend? Is it possible tweaking the code somewhere?

4

Hi @nihalsharma. If you want to view large number of records, you have to page the data. No viewer will be able to load millions of records in one go.
As I understand, redash will load the results and then page the table visualization. But this means all the data is already in memory before it can be paged.
I would page the data before it reaches redash. The remaining issue is finding a way to let the user select the page to load.
I have not done this but I imagine you could have a parameter to this query that simulates paging?