For rough estimate, I’ve seen the following pattern.

Sudo code
Do queries by token range at random.

Select asjson * from table;

Take the length of json string of each row.

Perform average.

Cheers.

From: Ayub M <[email protected]>
Reply-To: "[email protected]" <[email protected]>
Date: Wednesday, December 11, 2019 at 11:17 PM
To: "[email protected]" <[email protected]>
Subject: average row size in a cassandra table

How to find average row size of a table in cassandra? I am not looking for 
partition size (which can be found from nodetool tablehistograms), since a 
partition can have many rows. I am looking for row size.

Reply via email to