Archive for May, 2008

Strange Ring Found Circling Dead Star

May 30, 2008

The headline published by NASA Science News was “Strange Ring Found Circling Dead Star,” a headline practically designed to set any true scifi fan’s heart a-flutter. Could it be? Wouldn’t it be wonderful if? Is it a…Ringworld?

Alas, it was not to be. The ring is gaseous and, while not explained, not nearly so interesting as Larry Niven‘s ring.


High Performance Java

May 30, 2008

I’ve been exchanging email about Java performance with Paul Hohensee, Sun’s technical lead on the Java SE VM. He sent me several slide decks and additional information that will be useful to anyone interested in writing high performance Java code or in extracting high performance from existing Java code.

The first presentation [PDF] contains a list of pointers to whitepapers on Java performance, tuning, and garbage collection, as well as additional resources on monitoring and management, including information on using Dtrace with Java. It also includes a list of performance-oriented blogs, which I reproduce here for your convenience:

Blogger Blog
Steve Bohne Steve Bohne’s Weblog
Dave Dagastine David Dagastine’s Weblog
Dave Dice Dave Dice’s Weblog
David Holmes David Holmes’ Weblog
Steve Goldman Travels with FatCatAir
Jon Masamitsu Jon Masamitsu’s Weblog
Keith McGuigan Keith McGuigan’s Weblog
Tony Printezis Tony’s Blog
John Rose John Rose @ Sun

The second presentation, titled “High Performance Java Technology in a Multi-core World” [PDF] by Paul Hohensee and David Dagastine was presented at the 2007 JavaOne Conference. It gives an overview of Sun’s multi-core architectures and then presents Java VM technologies and software optimization techniques to maximize application performance.

Paul also pointed me to this recent article on Java performance myths, which summarizes the conclusions of two Google engineers who took a close look at conventional wisdom versus reality.

A Good Week for Operating Systems

May 30, 2008

I’ve had a good desktop OS week.

First, I installed the first official release of OpenSolaris on my Apple MacBook Pro, which went flawlessly. It was better than the earlier release candidates and, for those only familiar with Solaris 10 installations, it is an entirely better, more modern and more enjoyable experience.

If you aren’t sure you will use OpenSolaris, but are curious about it, then download the LiveCD image and play with OpenSolaris in memory with no commitment to write it to your local disk. Get the OpenSolaris 2008.05 LiveCD image here.

I opted to use OpenSolaris within VirtualBox on my Mac rather than create a separate Bootcamp partition for it. In the past, I’ve used both VMware Fusion and Parallels Desktop for my virtualized OS instances, but I’ve switched to VirtualBox. Partly because Sun now owns the technology, but also because it is free. I’m running V1.6 and have had no problems using it with OpenSolaris. Get VirtualBox here.

This was also the week that Apple finally released 10.5.3, a much-needed Leopard update that fixes lots of bugs. In my case I’ve been able to re-enable turn power management and allow my MBP to sleep again–it now seems to be working properly. Even better, closed-lid mode works again and I am typing this entry using my laptop with an external USB keyboard and the Sun 24″ LCD monitor in my office. Apple’s Vista versus Leopard TV advertisements are now funny again.

I have indeed had a good desktop OS week.

Idiots with a Voice: The Paisley Plot

May 29, 2008

Apparently, black and white scarves are “out” this year. You can wear one, but beware being labeled a terrorist or a supporter of terrorists by paranoid “pundits” with apparently nothing better to do than spew absurd crap like this.

Even if we allow that what looks like a paisley scarf to me is, in fact, a kiffiyeh, so what? Millions of people wear these every day. And millions of them are not terrorists.

I’m pretty sure Sirhan Sirhan was wearing pants when he shot Robert Kennedy. I want to assure the practitioners of political correctness that I in no way support the idea of assassination by wearing pants.

President Bush Signs GINA!

May 22, 2008

With President Bush’s signature yesterday, the Genetic Nondiscrimination Act (GINA) has become law. This is excellent news for anyone considering genetic testing who might justifiably be worried about their future eligibility for health insurance being effected by the results of their testing.

Let me illustrate by personal example. My family has a history of Hypertrophic Cardiomyopathy (HCM) which causes thickening in the central wall of the heart and which can lead to complications, including sudden death. HCM has been linked to mutations in genes related to the formation of heart muscle. There are now genetic tests available that have a reasonably good chance of finding whether a person carries known mutations related to HCM. To use the test, one family member with known HCM is tested to see if any of the known HCM markers are present. If markers are found, it is easy and inexpensive to test additional family members for HCM. This is where GINA becomes important.

Because of the risk of sudden death with HCM and because HCM may not be initially detectable in children, using such a genetic test on my nieces and nephews would allow their parents to know for sure whether any of the kids have HCM or not and to take appropriate precautions. The problem has been, however, that if a child is found to have the mutations related to HCM, this may be considered a pre-existing condition by insurance companies later in the child’s life which may affect their ability to obtain insurance. GINA improves the situation by making it unlawful for insurance companies to discriminate against people based on the results of genetic testing.

Full details on GINA are here.

One of My Worst Trips to the West Coast

May 21, 2008

I flew from Providence, Rhode Island to San Francisco yesterday through Charlotte, North Carolina on US Air. I left my house at 10:30 am and was still in transit AND STILL ON THE EAST COAST eleven hours later. I eventually arrived at my hotel in Menlo Park at about 1:30am Pacific time, an amazing 18 hours later.

I flew from Providence to Charlotte without incident–everything went sucky in Charlotte. First, US Air reset the departure time to San Francisco several times. Eventually they decided due to brake problems that we would change both gates and planes. So an Airbus worth of passengers dutifully trudged the entire length of the airport to the new gate. After more delays, we boarded the 2nd plane. Over an hour later we were still waiting because a problem had been found in one of the fire suppression systems…a problem they apparently had trouble isolating and fixing.

Around 7:30pm Eastern, about three hours after our originally planned departure, they decided it would take at least another hour to diagnose the problem so they unloaded us and distributed $10 meal vouchers to everyone. To add to the fun, heavy rains and lightning had started so no planes were leaving in any case.

About 30 minutes later they then located a third plane for us to try—back at the original gate a the other end of the airport. So, more trudging. More waiting. We finally took off at about 9:30pm Eastern, landing at 11:45pm Pacific at SFO. I made it to my hotel sometime after 1am PT.

I opted for this indirect flight over a direct BOS->SFO flight to use up a $500 airline credit before it expired, something I thought would be good to do. It clearly ended up not being worthwhile given the amount of personal wear and tear this trip caused. Even if the delays had not occurred, it isn’t clear to me that flying down the east coast for two hours and then sitting in an airport for two hours before flying cross country makes any sense at all for either me or for Sun.

I am so looking forward to my flight home. Yep, complete with another lay-over in Charlotte.

Growing Flowers with Datacenter Heat

May 14, 2008

The Open Source Grid and Cluster Conference is being held this week in Oakland, California. I attended the first day of the conference before flying home to meet a personal commitment. My favorite talk of the day was Paul Brenner‘s presentation titled Grid Heating: Dynamic Thermal Allocation via Grid Engine Tools.

Brenner, who works as a scientist in the University of Notre Dame’s Center for Research Computing, is exploring innovative ways to exploit the waste heat generated by HPC and other datacenters via partnerships with various municipal entities in the South Bend area. His first prototype, currently in progress, involves placing a rack of HPC compute nodes at a local municipal greenhouse, the South Bend Greenhouse and Botanical Garden.

The greenhouse had recently been forced to close portion of its facility due to high natural gas heating costs. Brenner wondered if he could help. Since current datacenters can be viewed as massive electricity-to-heat converters (with a computational byproduct), it seemed there might be an opportunity to exploit the waste heat in some useful way. But transferring heat, especially low-grade waste heat, over distances is very inefficient. Was there a way to overcome this barrier?

Enter grid computing with its ability to harness remotely located compute resources. If Brenner couldn’t transport the heat to the greenhouse, why not place the datacenter at the greenhouse? The garden gets the heat and Notre Dame gets the compute resources via established grid computing capabilities like Sun’s Grid Engine distributed resource manager, which is already in use at Notre Dame. Cool idea? Hot idea!

Based on early prototype work which involves placing single rack in the greenhouse, the idea looks like a promising way to reduce natural gas heating requirements for the facility. Brenner has shown he can use grid scheduling software to deliver a desired temperature (within a range, of course) by simply adding or throttling compute jobs on the greenhouse cluster, which communicates with Notre Dame via a wide-area wireless broadband connection.

He has looked at humidity issues and so far they don’t seem to be a problem given the ranges supported by typical compute gear. And he points out that while the greenhouse environment does not offer the highly filtered environment of a controlled datacenter, the particulate tolerance for typical compute gear is far in excess of EPA guidelines for people.

Phase II will involve placing three full racks of gear at the greenhouse to significantly reduce heating costs. Notre Dame will pay the electrical costs and use the compute resources. The city saves money on heating.

While the greenhouse is an interesting experiment, it is not ideal since its heating requirements will fluctuate seasonally. There are, however, other installations that have constant heating requirements–for example, hospitals have a 24×7 need for hot water. Sites like this could be interesting for future deployments.

Brenner’s full presentation is available [PDF].

Now THIS is Peachy!

May 5, 2008

I’m a bit late posting this, but did want to mention that the Peach open movie project recently released Big Buck Bunny, a 3D animated movie that was rendered on the Sun Grid Compute Utility at Network.com. Details on the operation of the Peach render farm are here. You can also click on the diagram below for a closer look at the overall IT setup for the project.

OpenSolaris binary distro now available

May 5, 2008

Today Sun announced the availability of our first OpenSolaris binary distro, OpenSolaris 2008.05, which is built from the OpenSolaris Project‘s open-source code base. You can download the LiveCD bits and read more at http://www.opensolaris.com. With LiveCD you can boot the distro without writing anything to disk and decide later whether you want to install it or not.

I’ve filed this under my HPC category. Why? Because future versions of this distro will form the substrate on which we intend to build a full HPC distro that will include both the OS and layered products and will address both application development and deployment. Components will include compilers, MPI library, developer tools, distributed resource management, as well as provisioning, management, and monitoring capabilities.

The definition and creation of this HPC distro will be run as an OpenSolaris project, which we will be starting in earnest soon. We invite any interested party to join the HPC Developer Community to get involved.