Things to think about for query performance

Let’s start with partition bucketing. This clever technique is all about slicing and dicing your data into digestible pieces to minimize data scanned. Imagine cutting your data into parts by range, then dividing these parts into tablets using a hash.

Next up, we have sorting columns, these bad boys can dramatically enhance your query efficiency. Let’s say, your SQL query often uses ‘site_id’ and ‘city_code’ as filter conditions, why not use them as sorting columns? Makes sense, right?

Also enable the query cache. Typically it’s not enabled by default. In certain situations, you can get a 10X in performance for your queries.

I have enabled query cache in 3.1 version of SR but it doesn’t work.
I have set enable_query_cache = true on FE node and restarted FE but don’t see any difference. Is there a way i can confirm this is enabled on my cluster?