Date: 2020-06-09 11:18:20 +0200
From: xavier <>
To: SQL devs <>
Version: 11.37.7 (Jun2020)
CC: @PedroTadim

Last updated: 2020-07-27 09:30:13 +0200

Comment 27800

Date: 2020-06-09 11:18:20 +0200
From: xavier <>

User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.106 Safari/537.36
Build Identifier:

Using 200 million pieces of data for query, the left fuzzy query takes longer than other query statements.
SQL 1：select * from t1 where id like '%584049';
SQL 2: select * from t1 where id like '%170711%';
For example, full fuzzy query(SQL 2) returns a result set of 5000+, which takes 386ms,.However, it takes 915ms for the left fuzzy query to return 59 results.

Reproducible: Always

Steps to Reproduce:

1.select * from t1 where id like '%584049';

Actual Results:

The concurrent execution of left fuzzy query performance is more obvious.

Is there any way to improve the performance of left fuzzy queries?

Comment 27803

Date: 2020-06-09 11:46:42 +0200
From: @PedroTadim

Hello xavier,

We have been fixing the optimization of simple like expressions this past week. Expressions such as '%584049' will use a faster route on Jun2020-SP1 release.

Best regards,
Pedro

Comment 27809

Date: 2020-06-10 03:32:41 +0200
From: xavier <>

Thanks.
Looking forward to the new version.

Best wishes，
Xavier.