Re: What you get is what you want
Jean Rodrigues <[email protected]> Mon, 15 Nov 2004 20:34:23 -0300 (ART)
| Newsgroups | gmane.comp.java.seda.user |
|---|---|
| Message-ID | <[email protected]> |
--0-1062165025-1100561663=:2060 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I'm getting nothing what you meant ... ;-) Why would we need a 'blocking enqueue' feature? I just can't realize the needs of it ... And, by the way, it should defeat the main proposal of SEDA which is to be scalable (everything that blocks does not scale). Regards, Jean Rodrigues Jean Morissette <[email protected]> wrote: > And a stage or a queue should only expect an API, not classes, to be provided in its environment. > At best, the thread class should implements an interface, so that you would use > > ((ISandstormThread)Thread.currentThread()).setBlocked(+/-) I agree with you :-) But all that is only implementation details. A more important question is the algorithm that we will use to decide if yes or not we allocate a new thread in a pool when other threads in this pool are already blocked in a blocking_enqueue operation. Here are some simples (and maybe wrongs) possibilities to decide to not allocate a thread: *One* thread in the ThreadPool wait to have the 'blocking_enqueue' lock (because another thread already have it). *All* threads in a ThreadPool wait to have the 'blocking_enqueue' lock. We want to enqueue (with blocking_enqueue) to a queue that is full, or almost full. Suggestions? Jean ------------------------------------------------------- This SF.Net email is sponsored by: InterSystems CACHE FREE OODBMS DOWNLOAD - A multidimensional database that combines robust object and relational technologies, making it a perfect match for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8 _______________________________________________ SEDA-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/seda-users --------------------------------- Yahoo! Acesso Grátis - Internet rápida e grátis. Instale o discador agora! --0-1062165025-1100561663=:2060 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit <DIV>I'm getting nothing what you meant ... ;-)</DIV> <DIV>Why would we need a 'blocking enqueue' feature?</DIV> <DIV> </DIV> <DIV>I just can't realize the needs of it ...</DIV> <DIV>And, by the way, it should defeat the main proposal of SEDA which is to be scalable (everything that blocks does not scale).</DIV> <DIV> </DIV> <DIV>Regards,</DIV> <DIV>Jean Rodrigues</DIV> <DIV><BR><BR><B><I>Jean Morissette <[email protected]></I></B> wrote:</DIV> <BLOCKQUOTE class=replbq style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #1010ff 2px solid"><BR>> And a stage or a queue should only expect an API, not classes, to be provided in its environment.<BR>> At best, the thread class should implements an interface, so that you would use<BR>> <BR>> ((ISandstormThread)Thread.currentThread()).setBlocked(+/-)<BR><BR>I agree with you :-)<BR><BR>But all that is only implementation details. A more important question <BR>is the algorithm that we will use to decide if yes or not we allocate a <BR>new thread in a pool when other threads in this pool are already blocked <BR>in a blocking_enqueue operation.<BR><BR>Here are some simples (and maybe wrongs) possibilities to decide to not <BR>allocate a thread:<BR><BR>*One* thread in the ThreadPool wait to have the 'blocking_enqueue' lock <BR>(because another thread already have it).<BR><B R>*All* threads in a ThreadPool wait to have the 'blocking_enqueue' lock.<BR><BR>We want t o enqueue (with blocking_enqueue) to a queue that is full, or <BR>almost full.<BR><BR>Suggestions?<BR><BR>Jean<BR><BR><BR>-------------------------------------------------------<BR>This SF.Net email is sponsored by: InterSystems CACHE<BR>FREE OODBMS DOWNLOAD - A multidimensional database that combines<BR>robust object and relational technologies, making it a perfect match<BR>for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8<BR>_______________________________________________<BR>SEDA-users mailing list<BR>[email protected]<BR>https://lists.sourceforge.net/lists/listinfo/seda-users<BR></BLOCKQUOTE><p> <hr size=1><a href="http://br.rd.yahoo.com/mail/taglines/*http://br.acesso.yahoo.com/">Yahoo! Acesso Grátis</a> - Internet rápida e grátis. Instale o discador agora! --0-1062165025-1100561663=:2060-- ------------------------------------------------------- This SF.Net email is sponsored by: InterSystems CACHE FREE OODBMS DOWNLOAD - A multidimensional database that combines robust object and relational technologies, making it a perfect match for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8