High performance computing startup Scali rolled out a high availability extension for its HPC management platform, called Scali Manage/HA. The new extension plugs into Scali’s Linux cluster management software, Scali Manage 5.4. It’s designed for fault tolerance on the Scali Manage server, providing access to the management console, configuration states and images in the event of a failure. It also provides access to cluster gateways enabling access to all nodes within clusters. You can check out Scali’s customer list for more info on the company.

I was just talking to Robert Sullivan, a senior consultant with The Uptime Institute, and he had some interesting things to say about the study on raised-floor air leakage that I wrote about yesterday. Sullivan is better known as “Dr. Bob” because of his PhD in applied mechanics, but more so for his expertise in data center physical infrastructure. Dr. Bob pretty much came up with the idea of a hot-aisle/cold-aisle design back in the early 90s.

Anyway, Dr. Bob was saying that air leakage depends a lot on how the raised floor is built. There are three situations:

1. A “stringerless” floor where tiles sit on pedestals, resulting in relatively large gaps between them and therefore a lot of leaking air.
2. A floor with pedestals and frames so that the tiles sit on the frames. Leakage is less here, but air can still get out by making its way between the frame and the tiles.
3. A floor with pedestals and frames, but also with rubber gaskets under each tile, which Sullivan said allows them to fit more snugly onto the frame, minimizing air leakage.

“You have to understand what the floor construction is to determine the relevance factor,” he said, adding that since many data centers use the rubber-gasket approach, the study could be “much ado about nothing.”

A server glitch caused the Dow Jones industrial average to drop about 200 points in less than a second yesterday. According to the Associated Press:

Dow Jones & Co., the media company that manages the well-known index of 30 blue chip stocks, said it discovered shortly before 2 p.m. that its computers weren’t properly handling the day’s huge volume in trades at the New York Stock Exchange.

It switched to a backup computer, and the result was a massive swoon in the index as the secondary system took over processing shortly before 3 p.m.

It looks like the plunge was caused by a backup of orders that suddenly registered when the backup server took over. Kenneth Polcari, a longtime stock trader, told Bloomberg: “The technology issues created some anxiety, which exacerbated the market’s move.”

You would think the Dow Jones would know how to handle a surge in traffic without messing up the industrial average, but apparently even they struggle with high availability.

Most data center power supplies are only 65-75% efficient. That means for every 100 watts you pump into that power supply, only three-quarters of it is actually getting to your server components. Talk about a huge waste of energy.

ColdWatt, an Austin, Texas-based company, claims that its power supplies are regularly 90% efficient, largely because of its patents in advanced magnetics technology.

The company is now trying to promote its products more broadly so IT workers will start demanding more efficient power supplies from server manufacturers. They’re not alone. There’s a group called 80 Plus that promotes power supplies that are 80% efficient or better by offering incentives to businesses that use them.

The Green Grid, a group formed by major IT vendors to foster environmentally friendly practices in the data center, announced that it has formed its charter and board of directors. It’s been 10 months since The Green Grid originally announced its formation. Members include AMD, APC, Dell, HP, IBM, Intel, Microsoft, Rackable Systems, SprayCool, Sun Microsystems, and VMware.

The press release also says there are three new white papers, but there really isn’t much to this announcement. Maybe the parties will start collaborating on data center efficiency and get end users involved. In the meantime, it’s still a wait-and-see.

Of the three white papers, the one on data center metrics (a PDF file) is probably the most useful for data center managers right now.

According to a recent report from Ashlee Vance at The Register, two Solaris execs have jumped ship from Sun Microsystems, including Solaris VP, Tom Goguen. Goguen had briefly served as a Solaris expert on SearchDataCenter.com. This could explain the delay in Goguen’s responses to reader questions. Goguen extolled the virtues of Solaris in a recent interview.

According to IDC, HP has held the top spot in the server market for the past 4.5 years, followed by Dell.

Users have their reasons for choosing servers. Some prefer Dell because they are pre-assembled and ready to use (ideally) when they arrive at your data center. They are also less expensive than the competition in most cases.

HP has a  reputation for quality hardware, and good tech support.

I’d like to know which you prefer and why.

Next week I’m headed out to the Pacific Northwest National Laboratory in Richland, Washington. I’ll be crossing the snowy Cascades, into Eastern Oregon and finally crossing the Columbia (The data center Promised Land) into Washington. It’s about a 6 hour drive from Eugene, OR but it’s going to be worth it.

 I’m going to be meeting with scientists and IT staff to talk about high performance computing, energy efficiency in the data center, and the projects underway at PNNL in the healthcare and airline sector.