Re: Sneer code, Async I/O and Vanilla-java

This is the operation context:

http://anonsvn.jboss.org/repos/hornetq/branches/Branch_2_2_AS7/src/main/org/hornetq/core/persistence/impl/journal/OperationContextImpl.java


The OperationContext is implementing the callback I use on the Journal

http://anonsvn.jboss.org/repos/hornetq/branches/Branch_2_2_AS7/src/main/org/hornetq/core/journal/IOCompletion.java
http://anonsvn.jboss.org/repos/hornetq/branches/Branch_2_2_AS7/src/main/org/hornetq/core/asyncio/AIOCallback.java


Any time either NIO or the native AIO SequentialFiles complete a sync
they will call the callback there.


Before I submit anything to the Journal, I call storeLineUp.

I then add something similar to Runnables to the Context by calling
executeOnCompletion(final IOAsyncTask completion)

One caveat I have is if the context is already complete at the time I
called executeOnCompletion, I simply call it at the same thread to
avoid context switch when not needed (say when the persistence was
faster).

With this context, I never blocked anything on the server and I could
scale up much better.

These are thing that I discovered over hard time on tests and during
implementation. Somewhat our secret sauce on the performance :-)


Feel free to ask me any help (even directly if you need). I may not
have a lot of time to dedicate but I somewhat have some love for
prevalyer. As I first heard about this project through Einar Saukas
and Bruno Souza when I was working a Summa in Brazil. I guess they are
friends in common from many of you guys who wrote prevalyer initially.

------------------------------------------------------------------------------
BlackBerry® DevCon Americas, Oct. 18-20, San Francisco, CA
Learn about the latest advances in developing for the 
BlackBerry® mobile platform with sessions, labs & more.
See new tools and technologies. Register for BlackBerry® DevCon today!
http://p.sf.net/sfu/rim-devcon-copy1 
_______________________________________________
To unsubscribe go to the end of this page: http://lists.sourceforge.net/lists/listinfo/prevayler-discussion
_______________________________________________
"Databases in Memoriam" -- http://www.prevayler.org