Re: [Quickfix-developers] QuickFIX performance

[Erich <gra...@al...>]

I've tried these and it seems not specifying the log factory makes a big
difference. It's much faster now! Many Thanks.


Djalma Rosa dos Santos Filho wrote:
> 
> QuickFIX Documentation:
> http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX Support: http://www.quickfixengine.org/services.html
> 
> Hi Erich,
> 
>>From my experience I believe that the decisions bellow will affect the
> QF performance.
> 
> 1. Selection of the proper Acceptor/Initiator (ThreadedSocketAcceptor
> or SocketAcceptor?)
> 2. The LogFactory (FileLog, MySQLLog, ODBC... or simply don't use the Log)
> 3. The StoreFactory (MemoryStore, FileStore, MySQL...)
> 4. It is nice to have the MessageCracker, but I think it is not cheap
> 
> - You should also consider some settings in configuration file:
> 
> PersistMessages=Y/N (should use N if you feel you don't need
> retransmissions)
> SocketNodelay=Y/N (TCP delay, in my app I had better results without the
> delay)
> DataDictionary=Y/N (to disable message validation, but without it QF
> cannot parse repeating groups)
> 
> I think that It would be nice to have settings in QF for TCP Window
> Size definition (SO_SNDBUF and SO_RCVBUF socket options), but Ok you
> can try different sizes by changing the OS settings directly and see
> how the application reacts.
> 
> 
> Djalma
> 
> On 9/10/07, Erich <gra...@al...> wrote:
>> QuickFIX Documentation:
>> http://www.quickfixengine.org/quickfix/doc/html/index.html
>> QuickFIX Support: http://www.quickfixengine.org/services.html
>>
>>
>> I am currently evaluating QuickFIX to compare its performance against
>> other
>> FIX Engines.
>> If I am interested in optimising QuickFIX to gain maximum throughput
>> under
>> load what configration settings should I look at? Are there any coding
>> pitfalls to be aware of?
>>
>> I am writing a C++ app, using FIX 4.2 and looking at order-related
>> messages
>> (e.g. New Order Single, Execution Report).
>>
>> Thanks
>> --
>> View this message in context:
>> http://www.nabble.com/QuickFIX-performance-tf4416685.html#a12597480
>> Sent from the QuickFIX - Dev mailing list archive at Nabble.com.
>>
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft
>> Defy all challenges. Microsoft(R) Visual Studio 2005.
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Quickfix-developers mailing list
>> Qui...@li...
>> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>>
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
> 
> 

-- 
View this message in context: http://www.nabble.com/QuickFIX-performance-tf4416685.html#a12615464
Sent from the QuickFIX - Dev mailing list archive at Nabble.com.