Monday, February 25, 2008

Adobe AIR. The More Things Change?

Erm ... Call me old fashioned but, apart from the implementation details such as AJAX, Flash, and yadda, yadda, isn't Adobe AIR like totally how the user interface worked prior to Netscape? Say, America Online circa 1992? I think it was called a client-side application. What am I missing here? Oh, I get it! Web 2.0 eliminates the marketing compulsion to put CDs in your airline food.

Paravirtualization in VMWare Server

Guerrilla alumnus Peter Lauterbach provides some updates on improved performance transparency in VMware ESX Server 3.5 (the bare metal virtualization product).

"Paravirtualization has been available in the VMware Workstation product since version 6, but it also requires guest kernel changes. Linux kernel 2.6.20 and above support pv, look for CONFIG_VMI=y in kernel config. The latest Ubuntu release has this enabled by default. You also need to check the paravirtualized kernel flag in the VMware Workstation and ESX Server Advanced settings under the Options tab.

Paravirtualization in Linux has quite a few moving parts to it, and affects mostly performance metrics and timekeeping functions. The best example is %steal time, which is the time that the virtual guest wanted to run, but was not able to, usually due to contention with other VM guests. This time was previously only visible to the virtual host, but can now be seen by the guest. Paravirtualized guests do run faster, and certainly feel more responsive when using a Linux GUI like KDE on a virtual guest.

Other advances from VMware are things like "VMDesched", which is an experimental feature in VMware Tools that creates a thread that is charged all the involuntary preemption time from the physical CPU."


Thanks, Peter.

Thursday, February 21, 2008

Cool Crypto Attack Gives Cold Comfort

A team of academics at Princeton University, together with industry and independent researchers has demonstrated a new class of computer attacks that compromise the contents of supposedly secure memory systems, particularly in laptops. The attack exploits the relatively slow decay of memory bits (seconds to minutes) that often contain disk encryption keys. The usual decay rate can be slowed even further by spraying the RAM chips with the compressed gas in a typical can of anti-static spray. A video presentation and the research paper are available at the Princeton site.

Saturday, February 16, 2008

Web 2.0 Meets Error 33

Apparently Amazon's Elastic Cloud snapped yesterday and havoc rained down on a number of Web 2.0 sites. This is unfortunate because the same kind of technology was deployed very rapidly (elastically?), exactly one year ago, to help search for missing computer scientist and yachtsman, Jim Gray.

When I was at Xerox PARC, we had a term for this kind of failure mode: Error 33. Error 33 states that it is not a good idea for the success of your research project to be dependent on the possible failure of someone else's research project. This term was coined by the first Director of Xerox PARC, Dr. George Pake and the nomenclature is reminiscent of Catch 22.

Error 33 is an all too appropriate reminder that a lot of Web 2.0 technology, which is hyped as ready for prime-time, is really still in the R&D phase. It's probably only very annoying when SmugMug is off the air for several hours, but mission-critical services like banks and hospitals should approach with caution. Achieving higher reliability is only likely to come at a higher premium.

Board from the Back of the Bus (maybe not)

When boarding a tour bus, the driver often tells you to occupy seats at the back of the bus first. This protocol is assumed to be more efficient than filling seats from the front because people block the aisle while stowing their carry-on baggage and thereby impede the flow. So, back-filling is more efficient than front-filling but is it optimal? The same procedure is used by airlines that implement boarding by groups A, B, C, etc. Group A is usually seated at the rear of the aircraft. There are some variants with aircraft boarding (e.g., window seats before aisle seats) that also help to distribute the passenger weight more evenly. The question remains, however, is it optimal?

An astrophysicist recently decided to look into this question more carefully using Monte Carlo simulations and found some surprising results. He unexpectedly discovered that the common rear boarding procedure is actually the second worst procedure, since it is only slightly more efficient than boarding front-to-back! So surprised was he, that at first, he thought there was a bug in his code. Then it became apparent that there was something more subtle going on. MC sims showed that an optimal boarding method was for passengers to board 10 at a time in every other row, since loading luggage requires about two aisles of space. In this way, passengers are either stowing luggage or sitting in their seats, rather than waiting in the aisle, as they do in the other two protocols. Depending on the size of the aircraft, this boarding procedure produces a speed up of 5 to 10 times over the worst case. Of course, disembarking is still LIFO. :-)

I wonder if this result could have applications in the context of computer or network performance analysis and capacity planning?

Tuesday, February 12, 2008

Guerrilla Training Schedule: A Two Point Landing

After many trials and tribulations, the 2008 training schedule has now stabilized at our new location: Larkspur Landing in Pleasanton, which is only about a mile away from the previous hotel location. The stumbling block for us was not room rates but outrageous contract penalty clauses for class cancellations. Apparently, hotel franchises, which are going to hurt this year in the already obvious economic downturn (I don't care what you call it), have suddenly decided they are no longer in the service/hospitality industry but the gold mining business.

Monday, February 4, 2008

Guerrilla Training Schedule Updates

Because of necessary hotel changes, we are still in the process of organizing Guerrilla Training Schedule for 2008. The bad news is, we still haven't been able to finalize the hotel. The good news is, we are converging. With any luck, everything should be wrapped up this week. Unfortunately, these ongoing negotiations have delayed our normal advertising schedule and one casualty is the new Guerrilla Boot Camp class, which we now have had to move out to July. We apologize to those who will be inconvenienced.

The latest schedule updates can always be found on the Classes page.

Friday, February 1, 2008

Penryn Performance. The Envelope Please!

Having kept pretty much to its promised schedule for the Penryn microprocessor ...

Intel is now significantly ahead of the industry with the production of 45 nm parts using the new high-K dielectric materials. The claims for the new Hafnium oxide-metal gate technology included:

  • Approximatley 2x improvement in transistor density, for either
    smaller chip size or increased transistor count

  • Approximatley 30% reduction in transistor switching power

  • Better than 20% improvement in transistor switching speed or 5x
    reduction in source-drain leakage power

  • Better than 10x reduction in gate oxide leakage power