Re: Slow query performance

[Shawn Green]

Clearly you are searching on some combination of values to see if a row already exists. This combination (what defines a unique row) for you should be either be the columns in a UNIQUE() key or the PRIMARY KEY for that table.

Another option is to build a numeric hash of a string composed of that combination of values then search for that number in an index. Lots of applications use the MD5() function for this but longer and shorter on-demand hash values are available if you review the documentation.

Then, if you do choose to use the PK or UNIQUE path, you gain the option of using the alternative command INSERT...ON DUPLICATE KEY UPDATE... to skip trips to the server. It would contain both paths of your attempt (it creates a new row if one does not already exist, or modifies the existing row if one did exist)
https://dev.mysql.com/doc/refman/8.4/en/insert-on-duplicate.html