Re: Out-of-Memory Handling in Long-Running / Server Applications

On Mon, Apr 14, 2008 at 5:14 PM, Randall R Schulz <[email protected]> wrote:
> On Monday 14 April 2008 13:38, Attila Szegedi wrote:
>  > On 2008.04.14., at 22:08, Randall R Schulz wrote:
>  > > The access patterns are such that to achieve anything near
>  > > acceptable performance, the tree and the priority queue must remain
>  > > in the working
>  > > set.
>  >
>  > You could look into Terracotta. While it's primarily a JVM clustering
>  > solution, almost as a side-effect it gives you a >4GB disk backed
>  > virtual heap where your RAM based heap basically serves as a cache.
>
>  Interesting. I've become aware of Terracotta recently, but this aspect
>  never occurred to me.
>
>  While I can't see that it would be suitable for the prover's main loop
>  where the heuristically driven search of an incrementally built tree
>  occurs, it might be helpful for another design problem I've yet to
>  solve.
>  > Attila.
>  Randall Schulz

(http://www.terracotta.org/confluence/display/explore/What+Is+Terracotta)
"Terracotta is Network Attached Memory (NAM). NAM is best suited for
storing what we like to call scratch data. Scratch data is defined as
object oriented data that is critical to the execution of a series of
Java operations inside the JVM, but may not be critical once a
business transaction is complete."

Sounds like a possible match to me.

--
[email protected]

===================================
This list is hosted by DevelopMentor®  http://www.develop.com

View archives and manage your subscription(s) at http://discuss.develop.com