Re: Error when test on DBT2 Postgresql

On 11/08/2015 04:50 PM, Tim Chou wrote:

Really CCing list this time.

Hi Adrian,

Thank you for your reply.

I use git to clone the repository (git clone
http://git.code.sf.net/p/osdldbt/dbt2 osdldbt-dbt2).
DBT2's website I used is
http://sourceforge.net/p/osdldbt/dbt2/ci/master/tree

I said my performance is very low because I read some papers and their
performance can exceed one thousand of transactions per second. I don't
want to compare with other databases, but I believe that my performance
is too low because only 100+ transactions can be processed.

I also try to change the data dir to a ram disk. However, the
performance keeps similar.

Which version of DBT2 and PostgreSQL do you use to test the performance?

I have not actually used DBT2. With the information you provided I downloaded it and tried to get it to run here. So far, ultimately unsuccessful using DBT2 or DBT3. I have gotten as far as the load test(loading the data) but then it fails for one reason or another. This is against 9.4. Going to take a break, maybe someone with more experience using DBT can make suggestions to you.

If I have some wrong understandings, please let me know.

Thanks,
Tim

2015-11-08 17:20 GMT-05:00 Adrian Klaver <adrian.klaver@aklaver.com
<mailto:adrian.klaver@aklaver.com>>:

    On 11/08/2015 01:55 PM, Tim Chou wrote:

    CCing list.

        Hi Adrian,

        The version o Postgres is 9.4.5
        I download the DBT2 from their sourceforge website.


    What site would that be?:

    When I go here:

    http://osdldbt.sourceforge.net/

    all the tests up to dbt4 download the same tarball which is for dbt1
    and is from 2003.

    When I go here:

    http://sourceforge.net/projects/osdldbt/files/

    the latest version is given as:

    dbt3-1.10.tar.bz2


        I follow the README file to run the test. (First, create and
        load data.
        Second, run dbt2-run-workload).


    In the README for the above I have:

    "Since this kit is derived from a TPC benchmark specification, TPC
    Policy on
    fair-use must be observed:


    http://www.tpc.org/information/about/documentation/TPC_Policies_v5.13.htm#_Toc124863463

    See the QuickStart for a brief introduction on using the kit.
    "

    So you will need to provide exact information as what dbt you are
    using and where it is coming from.


        Thanks,
        Tim

        2015-11-08 9:45 GMT-05:00 Adrian Klaver
        <adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>

        <mailto:adrian.klaver@aklaver.com

        <mailto:adrian.klaver@aklaver.com>>>:


             On 11/07/2015 11:27 PM, Tim Chou wrote:

                 Hi All,

                 When I test the DBT2 with a large number of connections, I
                 always get
                 the error:

                 Error in read.table(file = file, header = header, sep =
        sep, quote =
                 quote,  :
                     no lines available in input
                 Calls: read.csv -> read.table

                 I have tracked the file and see the reason is there are no
                 "TERMINATED"
                 in mix.log.
                 However, in my observation, driver and client do not close
                 elegantly so
                 they cannot quit with sprintf with the "TERMINATED"
        message (These
                 threads are closed by killall command).

                 1. How can I solve the problem? (I have checked the
        bash script
                 and c
                 code. I don't know why each thread cannot exit from the
        main
                 loop after
                 the predefined time. It should exit the main loop at
        time xxxx,
                 which is
                 a pre-computed value, according to the code.)

                 2. I want to know if my performance is too low.
                 No matter how I change the parameters, like creating
        RAM disk,
                 improving
                 the connection number, I always get the similar results.
                 Here is one of my test results:
                 Hardware: 4-core, 16GB of memory, two disks, and 1Gb
        ethernet
                 Parameters: 10 warehouses, 30 connections.

                                            Response Time (s)
                    Transaction      %    Average :    90th %        Total
                    Rollbacks      %
                 ------------  -----  ---------------------  -----------
                 ---------------
                    -----
                       Delivery   3.00      0.012 :     0.018            7
                    0   0.00
                      New Order  41.63      0.006 :     0.008           97
                    0   0.00
                 Order Status   5.58      0.003 :     0.004           13
                        0
                     0.00
                        Payment  47.64      0.003 :     0.004          111
                    0   0.00
                    Stock Level   2.15      0.005 :     0.007            5
                    0   0.00
                 ------------  -----  ---------------------  -----------
                 ---------------
                    -----
                 118.78 new-order transactions per minute (NOTPM)
                 0.8 minute duration
                 0 total unknown errors
                 97.0 seconds(s) ramping up

                 It seems only 2 txns/sec. Is that too low? How can I
        improve the
                 performance?

                 3. How can I test the performance in distributed mode?

                 Looking forward to your replies.


             What version of Postgres?

             Where did get the DBT2 test?

             How did you run the test?


                 Thanks,
                 Tim



             --
             Adrian Klaver
        adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>

        <mailto:adrian.klaver@aklaver.com
        <mailto:adrian.klaver@aklaver.com>>




    --
    Adrian Klaver
    adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>

--
Adrian Klaver
adrian.klaver@aklaver.com