what's the difference between Join Inner Tuple Table VS Join Inner Tuple Buffer?
I am trying to find out which inner join takes most of the memory, and use the following query.
SELECT
transaction_id,
statement_id,
path_id,
counter_tag,
COUNT(*),
(MAX(counter_value)/1024/1024/1024)::decimal(18,2) max_mem_gb,
(MIN(counter_value)/1024/1024/1024)::decimal(18,2) min_mem_gb,
(AVG(counter_value)/1024/1024/1024)::decimal(18,2) avg_mem_gb,
(SUM(counter_value)/1024/1024/1024)::decimal(18,2) sum_mem_gb
FROM
v_monitor.execution_engine_profiles
WHERE
counter_name ='peak reserved rid memory (bytes)'
GROUP BY
transaction_id,
statement_id,
path_id,
counter_tag
HAVING
sum_mem_gb > 100
OR max_mem_gb > 5
and the counter tag shows these two name, but what's the difference between them? Should we sum it up when check the memory used? or use the max of them? 