From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00
	autolearn=unavailable autolearn_force=no version=3.4.4
Path: 
 eternal-september.org!reader01.eternal-september.org!reader02.eternal-september.org!news.eternal-september.org!news.eternal-september.org!feeder.eternal-september.org!aioe.org!.POSTED!not-for-mail
From: "Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>
Newsgroups: comp.lang.ada
Subject: Re: GNAT.Sockets Streaming inefficiency
Date: Thu, 8 Jun 2017 16:46:51 +0200
Organization: Aioe.org NNTP Server
Message-ID: <ohbo0o$7gs$1@gioia.aioe.org>
References: <c00a464f-5587-4a8e-a11a-589f7030127a@googlegroups.com>
NNTP-Posting-Host: vZYCW951TbFitc4GdEwQJg.user.gioia.aioe.org
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
X-Complaints-To: abuse@aioe.org
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
 Thunderbird/52.1.1
X-Notice: Filtered by postfilter v. 0.8.2
Content-Language: en-US
Xref: news.eternal-september.org comp.lang.ada:46925
Date: 2017-06-08T16:46:51+02:00
List-Id: <comp.lang.ada>

On 08/06/2017 12:36, masterglob@gmail.com wrote:
> Configuration: X64, Linux & Windows (GNATPRO 7.4.2)
> While using GNAT.Sockets.Stream_Access, there is a real performance issue while using String'Output(...).
> 
> My test sends 500 times a 1024-long String using String'Output(TCP_Stream on 127.0.0.1) and the result is:
> - Linux : average 'output duration = 3 us
> - Windows: average 'output duration = 250 us
> 
>  From prior discussion with AdaCore, the "String" type is the only one for which this latency is NOT observed on Linux.
> 
> Any idea on:
> - is there a way to get similar performance on Windows (maybe using another type or method?)
> - is there any configuration that may solve this issue?

As a rule you should never use socket stream, especially if you are 
concerned about latencies. Because then you should have used NO_DELAY 
option on the sockets at the both sides. You would send full protocol 
packets (which with NO_DELAY set become frames). Consequently you would 
use Send_Socket once the packet is in the output buffer 
(Stream_Element_Array).

P.S. If you use socket stream then do stream's Write operation.

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