comp.lang.ada
 help / color / mirror / Atom feed
From: "Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>
Subject: Re: Prepared_Statement :(
Date: Tue, 27 Nov 2018 19:00:29 +0100
Date: 2018-11-27T19:00:29+01:00	[thread overview]
Message-ID: <ptk0nt$1n1r$1@gioia.aioe.org> (raw)
In-Reply-To: 37a5643e-02aa-48ab-b327-f76771d1f6e6@googlegroups.com

On 2018-11-27 10:58, eduardsapotski@gmail.com wrote:
> Dmitry, if I wanted to get a universal solution, I would never use Ada-language.

I doubt universal solutions, especially if not in Ada! (:-))

> Over year ago I wrote trading terminal on Java. Uses JDBC and can easily change db-driver.

Surely they cannot. RDBMS are way too different for this to work. Any 
"universal" solution must include a DB adapter to provide a reasonable 
application interface instead of the SQL mess (prepared or uncooked (:-)).

> And application can be easily transferred to any server.
> But what's the problem?
> Year ago, trading was conducted on ten currency pairs, today more hundred pairs!
> Java slows down and eats resources.
> Server is running at the limit.
> I am not so rich to buy Oracle SPARC M8.
> Ada significantly reduces server load.

Though Ada is far faster than Java it alone does not guarantee 
performance if the DB is the bottleneck.

Ada allows very efficient custom persistent storage implementations when 
RDBMS fails. Here is a small comparison SQLite vs B-trees in Ada:

http://ada-programming.blogspot.com/2014/09/ada-direct-io-b-tree-vs-sqlite.html

 > For this reason, I am not going to use anything other but PostgreSQL.
 > PostgreSQL very fast on insert operations. Even much faster than 
Oracle, other things being equal.

I cannot tell if PostgreSQL is better than Oracle or MySQL or whatever. 
One should really benchmark them for the specific case to know.

What I know is that the difference cannot be any dramatic. If one of 
them stops working all others probably will, and conversely. So if you 
have a real DB problem you possibly must say goodbye to all relational 
SQL-based DBs and look for a better custom algorithmic solution.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de


  reply	other threads:[~2018-11-27 18:00 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-24 10:45 Prepared_Statement :( eduardsapotski
2018-11-24 10:48 ` eduardsapotski
2018-11-24 11:05 ` Dmitry A. Kazakov
2018-11-24 11:20   ` eduardsapotski
2018-11-24 17:56     ` briot.emmanuel
2018-11-24 20:00       ` Simon Wright
2018-11-24 20:45       ` Dmitry A. Kazakov
2018-11-26  6:52       ` eduardsapotski
2018-11-27  0:27         ` Dennis Lee Bieber
2018-11-27  8:46           ` Dmitry A. Kazakov
2018-11-27  8:51             ` briot.emmanuel
2018-11-24 11:29 ` eduardsapotski
2018-11-27  9:58 ` eduardsapotski
2018-11-27 18:00   ` Dmitry A. Kazakov [this message]
2018-11-27 19:02     ` Björn Lundin
2018-11-28  5:04 ` eduardsapotski
replies disabled

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox