Понедельник, 13 июля 2020, 18:33 +03:00 от Rajendra Singh Negi <harharharishwar@gmail.com>:
 

Hi,

I have a background in SAP IQ (Columnar DB), SAP ASE(OLTP / row DB), SAP SQL Anywhere(OLTP / row DB).

I am interested in MonetDB as I think it is a Columnar Database.

 

Here are some points: 

1. Having a background in Columnar DB (SAP IQ), I would assume that the behaviour of MonetDB would be similar.

   (a) I would expect Columnar Databases to be slow in row inserts.

   (b) I would expect Columnar databases to be faster in bulk loads (I will explore this in monetDB)

        For example SAP IQ 16.x is the Fastest Bulk loading database in the world (https://www.guinnessworldrecords.com/world-records/fastest-loading-of-big-data)

        I would expect a good performance in monetDB if it has a bulk load capability.

 

2. OLTP / row Databases like ASE, SQLA, Oracle, mySQL, mariaDB, etc should be better than Columnar databases in row-inserts  

 

Thanks,

 

Sincerely,

 

Rajendra Singh Negi

Mobile: 7709112263

email: harharharishwar@gmail.com

Chase the Dream and not the competition

 

On Mon, Jul 13, 2020 at 3:13 PM Martin Kersten <martin.kersten@cwi.nl> wrote:

Hi

Could you try out this version https://github.com/bolner/MonetDB-PHP-Deux
and report on your findings.

regards, Martin

On 13/07/2020 08:52, - - wrote:
> hi Martin,
>    monetdb version: 11.37.7-20200529
>    sql: generic insert — INSERT INTO table(id,v1,v2,v3,v4) VALUES (…);
>    test: same environment for both servers: vmbox 8gb, 4 cores, only one server is running on tests, no load of the host server
>    php 7.4.5,  SQLAnyhwere pdo extension;  php-monetdb-1.0-1.el8.noarch.rpm
>    script: same for both tests except connection, measured only loop execution,   the script uses «random» functions to
>    generate  data for insert
> function str_random($length = 3200)
> {
>      $pool = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ';
>      return substr(str_shuffle(str_repeat($pool, $length)), 0, $length);
> }
>    the execution time ratio for all tests was almost the same. each test was done on «clean» server.
>    one of the test:
>    monetdb:  Work: 796.90835905075 seconds
>    SQLA:      Work: 382.80270600319 seconds
>    /alex
>
>     Понедельник, 13 июля 2020, 8:44 +03:00 от Martin Kersten <martin@monetdb.org>:
>     Hi <who am i talking to?>
>
>     There can be many reasons. Ranging from version used, SQL difference, test running, en PHP? API.
>     Based on the information provided this can not be judged.
>
>     regards, Martin
>
>     On 13/07/2020 02:13, - - wrote:
>      > I have run the series of INSERT tests in PHP for monetdb and SAP SQLAnywhere.  The tests show: performance of the monetdb on INSERT operations is almost twice slower against SQLA with the equal conditions (DDL, index, data, autocommit, 10K records),
>      > monetdb is in in-memory mode . What can be a reason?
>      > p.s. It is not benchmark, I just used the SQLA test server already running.
>      > --
>      > - -
>      >
>      > _______________________________________________
>      > users-list mailing list
>      > users-list@monetdb.org </compose?To=users%2dlist@monetdb.org>
>      > https://www.monetdb.org/mailman/listinfo/users-list
>      >
>
>     _______________________________________________
>     users-list mailing list
>     users-list@monetdb.org </compose?To=users%2dlist@monetdb.org>
>     https://www.monetdb.org/mailman/listinfo/users-list
>
> --
> - -
>
> _______________________________________________
> users-list mailing list
> users-list@monetdb.org
> https://www.monetdb.org/mailman/listinfo/users-list
>
_______________________________________________
users-list mailing list
users-list@monetdb.org
https://www.monetdb.org/mailman/listinfo/users-list

_______________________________________________
users-list mailing list
users-list@monetdb.org
https://www.monetdb.org/mailman/listinfo/users-list