I was at TheServerSide.com’s Java Symposium last week and got a fascinating perspective on where high-end Java apps are headed, and the infrastructure that will be needed to support them.

Hedge funds and more traditional financial service firms all are deep into creating what they call grids (they’re not talking about time sharing across occasionally idle computers) for doing performance intensive stuff like programmed trading. Imagine hundreds of motherboards ripped out of servers and velcroed into racks, all running stripped down Linux cores and highly tuned Java Virtual Machines (JVMs) on top.

Kirk Pepperdine, veteran Java performance tuning consultant, discussed the growing reliance on non-volatile memory over disk (you’ll have to register to download the talk, but it’s pretty easy) to reduce latency in these applications. NVRAM can mean high-memory footprint motherboards, as we are now in the hundreds of gigabytes for some systems. But it can also mean solid state disk, which maybe undergoing one of its periodic surges. Not only are people trying to put entire programs in memory, but as much data as they can, too.

Typical databases of, say, 1990, would easily fit into today’s NVRAM. But not necessarily the databases of today, which have grown into millions and millions of rows. But objects and services give developers an option they are starting to explore: stuffing the data into the objects themselves and stuffing the objects into solid state caches, whether that’s onboard or outboard.

One thing that is making this possible is using the grid. Some shops are using what is effectively distributed memory, as Iona Technology’s technical director, John Davies, pointed out. Products like GiagSpaces, Oracle’s Coherence and GemFire create and manage a memory space across many machines.

Another take comes from Azul Compute Appliance, your classic black box.The company has attacked a specific and troublesome problem with Java apps – garbage collection. The JVM can pause for as long as 30 seconds every few minutes to do its thing. Even if garbage collection isn’t that extreme, brief pauses are not acceptable for high-performance trading apps. So Azul designed its own chips that use a proprietary instruction set to make garbage collection non-disruptive. They stuff up to 768 of them in a box, with up to 768GB RAM. Software on the hosts redirects calls to the JVM to the Azul box, where it runs as if it was on the host.

Azul’s boxes run in the tens and hundreds of thousands of dollars, but they can run many JVMs at once. According to benchmark’s run by Pepperdine, they definitely turn up the heat on Java apps.

Server managers in the big Wall Street firms are already dealing with these new concepts, and you can expect them to migrate outward in coming years. Just as storage and networking have been disaggregated from the computer, some amount of memory and processing, at least for specialized purposes, may also migrate on to the network (probably 10GigE or InfiniBand). One of the big reasons pointed out by Pepperdine: with multicore processors, clock speeds are not increasing. Therefore, app developers must seek other ways to increase performance.

Given that parallelism is still only minimally doable for all but the rocket scientists, techniques like greater use of caching are bound to gain popularity, so it’s probably worth your while to start investigating this whole technology area. At the architecture level, there’s plenty to understand: should you virtualize and cluster at the app level, or go the route of Virtual Iron or ScaleMP, which allow you to concatenate multiple physical machines into a single large VM?

Never a dull moment.

I saw this interesting story in the Indianapolis Star about a data center being built where a brick and mortar mall once stood.

According to the report, Indiana-based colocation facility Lifeline Data Centers agreed to buy Eastgate Consumer Mall and invest $23 million to renovate it and lease it out to companies for data storage.

Lifeline, with only eight employees, is seeing explosive growth in demand for its services. Co-owner Rich Banta, a former chief technology officer for Community Health Network in Indianapolis, said, “The sales pipeline is what’s driving this,” as computer needs for companies and public-sector employers continue to grow and take up more space,” the report said.

The center will employ about 1,500 information technology employees and will consume as much electricity as used by a town the size of Lebanon, Lifeline told the reporter.

To me, this paints the picture of where we are today: shopping is done online instead of in actual malls, where crowds annoy and danger looms, and mall space is being filled with servers to power those e-commerce sites.

Of course, this is one instance and I don’t know the circumstances of this mall closing, but it seems to be a reasonable analysis from this online shopper’s amateur perspective.

” There are very few places where a computer is so central that everything crashes to a halt if the machine goes on the blink.” This is a quote from a recent article from the New York Times about the possibility of cyberwar. The article is basically about how th U.S. and other countries, namely China, probe each other’s stability and computer infrastructure security to see who is vulnerable and where and how. The author points to cyberattacks in Estonia, where tensions with Russia prompted Russian ‘hacktivists‘ to jam critical web servers (i.e., banks, military, etc.) with so much digital debris that they basically shut down access from normal users, as an example of an impending cyberwar.

Perhaps it’s just my perspective from an IT journalist vantage, but I disagree with the assertion that a cyberattack would merely be an inconvenience for any nation with so much digital infrastructure. Especially as we continue to move to web-based services to improve change and configuration management (yes, for critical back-end stuff, we’re talking about a secure intranet, but it’s like anything - the more moving parts . . . ), I think that it is something that data center managers and server operators need to take seriously.

Reuters is reporting that Microsoft plans to release a “very, very large” number of web services that would allow smaller web developers to take advantage of Microsoft’s impressive data center network. Silverlight, Microsoft’s answer to Adobe’s Flash, is only the first in the software giant’s attempt to create relevant products that rely on more than name recognition and a hegemonic relationship with ‘computing’ and other tech-terminology. Will Silverlight be able to make up for coming to the game late, as their virtualization effort is hoping to do? My money is on no. MS has deep pockets, the most recognizable name in technology, and good products; but there are already conventions in Web encoding and services (read: YouTube) that people are happy enough with now.

SmugMug CEO Don MacAskill has amazing hardware reviews on his blog. If you’re looking for very candid feedback on how server vendors stack up at a growing Web 2.0 company, you need to read these posts.

According to MacAskill, Sun’s hardware engineering is awesome, but the sales cycle still sucks. Sounds familiar.

Here are the greatest hits:
SmugMug review of the SunFire x2200 M2 server: Two months after SmugMug dumped Rackable for Sun, MacAskill says he’s still starry-eyed in love. Pie in the sky, he’d like to see dual power supplies, DC-power servers and a liquid cooling option.

x86-64 Server vendor bakeoff: Server vendors vie for SmugMug’s newly minted Web 2.0 dollars. HP doesn’t fit the bill, passes them off to a VAR; Dell is quick on the draw but not paying attention to what they actually asked for; IBM doesn’t return phone calls (Tell me about it!); Rackable is “cheap in every sense of the word” and its once “cool” server business is in some sort of freefall. And the winner by default: Sun.

SmugMug details Amazon’s S3 performance flaws: Amazon’s storage offering, S3 had some major issues according to MacAskill. “Customers were at our gates with pitchforks and torches. Our paying customers were affected and they could tell there was something wrong. Not good.” SearchStorage.com also reported on customer problems with S3.

A review of Sun’s T1000: Why SmugMug didn’t go with Sun’s T1000 server.

Blog company Wordpress has beefed up its server farm and spread out into three data centers according to Barry at Wordpress. The company has spent the past month moving out of a facility in San Diego and moving into two facilities in Texas and one in San Fran. The company has standardized on HP’s rack machines, the bulk of which are AMD-based HP Proliant DL145 servers.

The new racks will also power Wordpress’ sister company, Askimet, a blog comment spam filtration tool. Interesting note: Askimet predicts that 95% of all blog comments are spam — I estimate that I delete a spam comment about every two hours.

Let’s say that you’re a data center manager for one of the many online companies competing for a piece of the digital media boom. As a recent article in the New York Times discusses, one of the primary obstacles between your company and financial success has to do with copyrights, licenses, things of that sort, for obtaining enough content to get traffic. Web 2.0-based sites, such as YouTube, Google Video and others, get quite a bit of their content from people who feel that their opinion is worth sharing with the rest of the world (yeah, I know I’m blogging), but still have to be legal-savvy about content they’re obtaining themselves as well as copyrighted UGC (user generated content). While all that “who owns the content” side of the business is complex and important, if not interesting, there are a few other parts worth thinking about.

Data centers need to be ready to handle the flux of information that is literally everywhere. Cell phones, laptops, PCs and other devices now capable of relaying content stored and accessed through a server are making the flow of information even more liquid. In other words, the ability to access video, music, images, etc. anytime and from anywhere, makes the web seem less like a web and more like an ocean where information seems like an organic mass. Given the number of options and the waning attention span of the next generation of users, downtime for your media servers could mean thousands if not millions of lost page views. A recent article by Beth Pariseau begins to approach this looming issue as she looks into user opinion on Amazon’s S3 tool. The sky isn’t falling yet (and it is likely that it never will), but at some point, there won’t be enough power and cooling, storage, etc. to keep up with the demand/generation of web content.

Also interesting, for me at any rate, is that the rush to obtain and host content such as live concerts, baby’s first words, and other “live” moments seems to reflect what theorist Jean Baudrillard said about simulacrum. We are creating an image of life that becomes life itself. Our digital media has become the copy with no original. The simulation is the real.