Richard - Thanks for your reply. I actually posted something on the discussion page of the Pluggable Storage wiki last August, but apparently it hasn't garnered many glances (http://wiki.dspace.org/index.php/Talk:PluggableStorage). I considered those same possibilities, but they all led to other cans of worms, particularly since we're expecting to see some really big files (well, too big for memory, anyway).
I'd be happy to have you look at my code. I have no experience in the open-source world, and I wasn't sure how I could contribute what I'd done in case anybody might be interested, but maybe this is how. What would be the best thing to do (after I give it the once-over to remove any obvious stupidisms and lazy hacks)? Send it to you? - Rick -----Original Message----- From: Richard Rodgers [mailto:[EMAIL PROTECTED] Sent: Friday, March 07, 2008 3:42 PM To: Rick Runyan Cc: [email protected] Subject: Re: [Dspace-tech] Pluggable Storage Question(s) Hi Rick: Pluggable Storage is definitely on near horizon, but it isn't part of the 1.5 distribution. I had been waiting for a few other adopters to insure that it was adequate. I believe we will see a Sun Honeycomb implementation soon. There are indeed issues with DB connection management - which is why I didn't add a RDBMS implementation at the outset. I'm not sure what the optimal solution is, but there might be possibilites in writing to a temporary memory or disk cache, where the usual read semantics would apply. If you want to send your code (or post to the wiki), I'd be glad to kibbitz... Thanks, Richard On Fri, 2008-03-07 at 14:50 -0500, Rick Runyan wrote: > Hi - > > > > I used the Pluggable Storage interface to modify a 1.4.2 test > installation to store bitstreams as blobs in an Oracle database. I'm > curious as to whether anybody has moved a working PS implementation to > 1.5. > > > > Also, about my implementation - I had to modify the interface so that > the bitstore methods' callers would pass in their Context. > > > > Strictly speaking, this was necessary only for the 'get' method. The > problem I had is that when those methods manage their own connections, > they have to close them before they return. That meant the connection > to the blob stream was closed before get's caller could begin reading > it. Has anybody encountered a similar problem? Am I overlooking some > painfully simple solution? > > > > - Rick Runyan > > > > > ------------------------------------------------------------------------ - > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ DSpace-tech mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dspace-tech ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ DSpace-tech mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dspace-tech
