Archive for the 'JBoss Cache' category

JBoss Cache clustering demo

In his session this morning on JBoss Cache, Manik Surtani demonstrated caching and clustering via the lightweight GUI described in “Visualizing JBoss Cache.” It’s a fairly straightforward GUI that gains from it’s simplicity. He mentioned that the GUI demo was available for download, which I found here.

In the demo, Manik is using a buddy node configuration. You can read more about buddy nodes and other clustering configurations via the links in “Full cache at JBoss World“. Regarding one semantic clarification from the talk, a node is a cache instance, that is, a tree node in the data structure. This does not refer to, for example, a physical server node. One audience member who asked about the usage gave the clear language of, “A node is an object that has state in our application and that state needs to be replicated.”


Full cache at JBoss World

First session of JBoss World is Bela Ban and Brian Stansberry talking about optimizing HTTP session clustering, focusing on JBoss Cache and JGroups. Their talk is full of immediately useful info, and I see lots of note taking. Fortunately for all of these attendees, and also for everyone ever after, this is also the first talk that I’m we’re capturing full audio video for Dev Fu technical session archives. That audio video plus the slides from the talk should may help you find the useful tips, such as this one on tuning and slimming JBoss Application Server by searching for the misspellings in the wiki name “tuning sliming jboss“.

This room is packed wall to wall, and more than a few on the floor. So far this is the sign for the whole conference. There are a lot of people here. Don’t know if there are ever official numbers given out, but it seems well into the hundreds and hundreds. Pretty good for a developer focused conference.

The Red Hat video story team is here in force. I wandered around for the first twenty minutes of getting here this morning, saying howdies to friends. My excitement didn’t kick in until I went into the little back room (in the middle of everything) where the video editing is taking place. Not that it was a lot of equipment or looking like the MI5 van in an episode of Spooks. It is a relatively modest nerve center, but it’s where I’ll be spending some afternoons and evenings and maybe early mornings all week long, working with this great crew to bring you the developer story happening here at JBoss World.

Update: When I greedily went to listen to the recording after the session, I found it blank! Curses. Apparently I pushed the [Rec] button just one time, which puts it in standby mode. It faithfully stood by for a while, then went to sleep. We’ll just have to see how the video comes out for giving us useful audio, but I don’t have a lot of hope. Oh, well, spilt milk and all that, no time to cry. Thanks to Bob McWhirter of JBoss.org for his video capture, which gives us something much more than nothing.


Visualizing JBoss Cache

When demonstrating the value of JBoss Cache, Manik Surtani deals with a common challenge — how to demonstrate what JBoss Cache does when it’s features don’t include a flashy GUI. Manik’s idea was to build a GUI demo that shows what JBoss Cache is doing under the covers. For example, one way to see functionality is to start multiple instances of the demo, watching them form a cluster and share data.

This reminds me of demonstrating other under-the-covers system services, such as virtualization or LVM and RAID. Even when it is lines of text flipping by, it is impressive to see services create, be destroyed, and recover gracefully.

This Friday I’m planning on attending Manik’s talk at JBoss World. It’s important to bring attention to projects that might not provide as much flash and bang as other tools, but are essential ingredients to a successful development and deployment. With that in mind, I’ve got Manik’s talk on my agenda of audio presentations I’m capturing, and we’ll discover just how flashy JBoss Cache is.