Odd Athena Query Performance Issue

0

Hello, I often use CTE's or views to create faux parameters as a way to improve reusability of code. For example, if I had a bunch of CTE's or subqueries (or different queries/views altogether) that all used a common value in a where clause, I would simply create a view/CTE containing the values and select the values from a subquery wherever needed. This typically has no meaningful impact on performance; however, it seems to degrade execution speed quite a bit in Athena and I'm not sure why this would be the case. What is the difference between: 'select * from table where foo = 'bar' versus 'with param as (select 'bar' as p) select * from table where foo = (select p from param)'? When querying a substantial amount of data, the second pattern makes the query speed untenable. These tables are partitioned to the hour also.

meowow
已提問 2 年前檢視次數 487 次
1 個回答
0

Hello,

Technically it should not effect the performance. Can you please provide Athena query ID's and AWS region for us to check the query plan and see where the delay is ? Feel free to open a support ticket with Athena support team with query ID's and we can provide you more insights.

AWS
支援工程師
已回答 2 年前

您尚未登入。 登入 去張貼答案。

一個好的回答可以清楚地回答問題並提供建設性的意見回饋,同時有助於提問者的專業成長。

回答問題指南