Hi all!

My benchmark consists of 1000 lines of identical queries with 3 seconds delay between them. Most queries returns the result in a short time (~ 6 ms) but some queries delay answer to 15000 ms. Than this may be due and how to check and eliminate this delay?


  • Options

    If you have the transaction_id, statement_id and initiator node of the slower query ( You could also get that info from the query_profile table) you could check where the query took time and depending on the phase it can be many things. The below query will show you the time in each of the phases,  

    select node_name,execution_step,completion_time-time as duration from dc_query_executions where transaction_id= <transaction_id> and statement_id = <statement_id> and node_name ilike '%<initiator_node> ' order by time;


    let us know where do you see the delay.



  • Options
    Thank you very much!
    I will analyze this table. I wonder why the same questions selected from a file, give a a wide scatter runtime.

Leave a Comment

BoldItalicStrikethroughOrdered listUnordered list
Align leftAlign centerAlign rightToggle HTML viewToggle full pageToggle lights
Drop image/file