Postgres Pro
Downloads
Home   >   mailing lists

Re: Postgres insert performance and storage requirement compared to Oracle - Mailing list pgsql-performance

From Divakar Singh
Subject Re: Postgres insert performance and storage requirement compared to Oracle
Date
Msg-id 25247.43880.qm@web65408.mail.ac4.yahoo.com
Whole thread Raw
In response to Re: Postgres insert performance and storage requirement compared to Oracle  (David Wilson <david.t.wilson@gmail.com>)
List pgsql-performance
Tree view




    for (; i<howmany;i++ )
    {

    sprintf (query, "INSERT INTO aaaa(a, b, c, d, e, f, g, h, j, k, l, m, n, p) VALUES (67, 'ABRAKADABRA', 'ABRAKADABRA', 'ABRAKADABRA', '1-Dec-2010', 'ABRAKADABRA', 'ABRAKADABRA',  'ABRAKADABRA', '1-Dec-2010', 99999, 99999, %d, 9999, 'ABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRAABRAKADABRA')", i);

    res = PQexec(conn, query);

      if (PQresultStatus(res) != PGRES_COMMAND_OK)
        {
          cout<<"error at iteration "<<i<<":"<<PQresultErrorMessage(res)<<endl;
                  PQclear(res);
          break;
        }
        //PQclear(res);


    }

>Why is that PQclear(res) commented out? You're leaking result status for every insert.


I did that purposely to see if cleanup part is contributing to any performance loss.
Right now in my test, memory leak is not a concern for me but performance is.
Though I understand that memory leak can also result in performance loss if leak is too much.
However, in this case, commenting or uncommenting this statement did not cause any change in performance.
 




pgsql-performance by date:

Previous
From: Tom Lane
Date:
Subject: Re: Regression: 8.3 2 seconds -> 8.4 100+ seconds
Next
From: Francisco Reyes
Date:
Subject: Re: Regression: 8.3 2 seconds -> 8.4 100+ seconds