MIT Research News' Journal
[Most Recent Entries]
[Calendar View]
Friday, January 24th, 2014
| Time |
Event |
| 5:00a |
Bringing the world reboot-less updates
It’s an annoyance for the individual computer user: You’ve updated your operating system, and now you need to reboot. This is so the computer can switch to the modified source code.
Imagine, however, having to update and reboot hundreds or thousands of computers operating in large companies and organizations: It can have a significant impact in lost time and money as computers and online services shut down, sometimes for hours. To avoid downtime, organizations will usually wait for low-traffic periods to update — but this can leave the servers outdated or vulnerable to cyber attacks.
In 2008, Jeff Arnold ’07, MEng ’08, along with a team of MIT computer scientists and engineers, began solving this issue by developing and commercializing software, called Ksplice, that automatically applies patches (security updates or bug fixes) to an operating system, on the fly, without requiring a reboot.
Based on Arnold’s award-winning MIT master’s thesis, the novel software compares changes between the old and updated code and implements those changes into a running Linux kernel — an operating system’s core data-processing component. In essence, it does something that could normally be achieved only by shutting down the operating system.
The software also incorporates novel techniques that remove the need for programmer intervention with the code (a trademark of performing updates without Ksplice), which decreases the cost and risk of error, Arnold says.
“The aim is to allow administrators the benefit of the update while eliminating both the cost and downtime for the users,” Arnold says.
After winning the 2009 MIT $100K Entrepreneurship Competition for the software, Arnold co-founded Ksplice, Inc. — with Waseem Daher ’07, MEng ’08, Tim Abbott ’07, SM ’08, and Anders Kaseorg ’08 — in Cambridge to launch it as a commercial product. Arnold served as the company’s CEO.
In just 18 months, Ksplice accumulated 700 customers — independent firms, government agencies, and Fortune 500 companies that were running the software on more than 100,000 servers.
Then, the startup sold for an undisclosed amount to technology giant Oracle, which is now providing the software to its Oracle Linux customers, which include banks, retail firms, and telecommunications companies worldwide. (After the purchase, the Ksplice team joined Oracle to help the company integrate the software in its products.)
As of today, Ksplice has only ever run on Linux operating systems. But Daher says the code is written in a way that should make it “potentially expandable to other products,” such as Mac and Windows operating systems.
Object focused
The process of updating running kernels is called “hot updating” or “hot patching,” and predates Ksplice. But Ksplice’s novelty is that it constructs hot patches using the object code — binary that a computer can understand — instead of the source code, computer instructions written and modified as text by a programmer (such as in C++ or Java).
Hot patching a program without Ksplice requires a programmer to construct replacement source code or manually inspect the code to create an update. Programmers might also need to resolve ambiguity in the code — say, choosing the correct location in computer memory when two or more software components have the same name.
Ksplice, however, hot patches the object code, using two novel techniques invented by Arnold. The first, called “pre-post differencing,” creates object code before a patch (“pre”) and object code modified by the patch (“post”) on the fly. It then compares the “pre” and “post” code to determine what code has been modified, extracts the changed code, and puts the code into its own updated object file, which it will plug into the running kernel. Essentially, it makes changes to functions modified by the patch, and points to relocated, updated versions of those functions.
The second technique, called “run-pre matching,” computes the address in computer memory of ambiguous code by using custom computation to compare the “pre” code with the finalized, running kernel (“run” code). For example, by going back and forth between the “pre” and “run” code, and examining the “pre” code’s metadata values, it gains enough information to identify the code that needs to be relocated to a spot in the running kernel.
This technique also doubles as a safety measure, as the process will abort the patch if it finds any unexpected object code differences between the “pre” code and “run” code.
Although today’s computer technology risks becoming obsolete within a few years, Ksplice (now 5 years old) is still a novel product, says Daher, who was Ksplice’s chief operations officer. “It’s still going strong,” he says.
The source of Ksplice
Ksplice’s roots trace back to 2006, when Arnold was charged with implementing a security update, for MIT’s Student Information Processing Board, that arrived on a weekday. Trying to avoid downtime while the servers were in heavy use, he delayed installing the update until the weekend. This wait, unfortunately, resulted in a cyber attack that required reinstalling all the system software.
“That’s what motivated me to think about this problem: You can’t bring servers down right away, and can’t wait until you have a chance to update, which can be a week, a month, or a year, in some companies,” Arnold says. “We needed better technology for solving this impossible problem for updating on the fly, as the system runs.”
Under the tutelage of Frans Kaashoek, the Charles A. Piper Professor of Computer Science and Engineering, Arnold started developing Ksplice for his graduate thesis, which earned second place for the Charles and Jennifer Johnson MEng Thesis Prize in 2008.
From there, grants and MIT’s $100K competition funded Ksplice, and Arnold gathered a team to help develop the technology. As the company grew, Arnold and Daher took the reins: They dealt with customers, sales, marketing, and accounting — “challenging for people with strictly computer science backgrounds,” Daher says.
For help, they turned to MIT’s Venture Mentoring Service (VMS), “which was helpful in mentoring, guiding us through applying for grants, and thinking about the business in the right way,” such as understanding the customers and the market, Daher says.
“Something that was key to our success was we had a good network of mentors and advisors, and many were part of the MIT community,” he says. “It was valuable to hear their war stories and get their take on some of the challenges they were facing.”
“MIT is very much part of the Ksplice story,” Arnold adds.
Arnold and Daher are now working on another software startup at the Cambridge Business Center — and still keep in touch with the VMS, they say. Being in the early stages, they can’t say much about the startup. “But we’re happy to be back building a startup,” Arnold says. | | 5:00a |
A new wrinkle in the control of waves
Flexible, layered materials textured with nanoscale wrinkles could provide a new way of controlling the wavelengths and distribution of waves, whether of sound or light. The new method, developed by researchers at MIT, could eventually find applications from nondestructive testing of materials to sound suppression, and could also provide new insights into soft biological systems and possibly lead to new diagnostic tools.
The findings are described in a paper published this week in the journal Physical Review Letters, written by MIT postdoc Stephan Rudykh and Mary Boyce, a former professor of mechanical engineering at MIT who is now dean of the Fu Foundation School of Engineering and Applied Science at Columbia University.
While materials’ properties are known to affect the propagation of light and sound, in most cases these properties are fixed when the material is made or grown, and are difficult to alter later. But in these layered materials, changing the properties — for example, to “tune” a material to filter out specific colors of light — can be as simple as stretching the flexible material.
“These effects are highly tunable, reversible, and controllable,” Rudykh says. “For example, we could change the color of the material, or potentially make it optically or acoustically invisible.”
The materials can be made through a layer-by-layer deposition process, refined by researchers at MIT and elsewhere, that can be controlled with high precision. The process allows the thickness of each layer to be determined to within a fraction of a wavelength of light. The material is then compressed, creating within it a series of precise wrinkles whose spacing can cause scattering of selected frequencies of waves (of either sound or light).
Surprisingly, Rudykh says, these effects work even in materials where the alternating layers have almost identical densities. “We can use polymers with very similar densities and still get the effect,” he says. “How waves propagate through a material, or not, depends on the microstructure, and we can control it,” he says.
By designing that microstructure to produce a desired set of effects, then altering those properties by deforming the material, “we can actually control these effects through external stimuli,” Rudykh says. “You can design a material that will wrinkle to a different wavelength and amplitude. If you know you want to control a particular range of frequencies, you can design it that way.”
The research, which is based on computer modeling, could also provide insights into the properties of natural biological materials, Rudykh says. “Understanding how the waves propagate through biological tissues could be useful for diagnostic techniques,” he says.
For example, current diagnostic techniques for certain cancers involve painful and invasive procedures. In principle, ultrasound could provide the same information noninvasively, but today’s ultrasound systems lack sufficient resolution. The new work with wrinkled materials could lead to more precise control of these ultrasound waves, and thus to systems with better resolution, Rudykh says.
The system could also be used for sound cloaking — an advanced form of noise cancellation in which outside sounds could be completely blocked from a certain volume of space rather than just a single spot, as in current noise-canceling headphones.
“The microstructure we start with is very simple,” Rudykh says, and is based on well-established, layer-by-layer manufacturing. “From this layered material, we can extend to more complicated microstructures, and get effects you could never get” from conventional materials. Ultimately, such systems could be used to control a variety of effects in the propagation of light, sound, and even heat.
George Fytas, professor of materials science and head of the polymer group at the University of Crete, Greece, says this is a "very novel idea, because it induces a directional phonic gap not existing in the layered structure." He adds that this finding "shows how well-established theoretical tools can predict new materials behavior, which is challenging for experimentalists."
The technology is being patented, and the researchers are already in discussions with companies about possible commercialization, Rudykh says.
The research was supported by the U.S. Army Research Office through the MIT Institute for Soldier Nanotechnologies. | | 10:00a |
Researchers develop new method to control nanoscale diamond sensors
Diamonds may be a girl’s best friend, but they could also one day help us understand how the brain processes information, thanks to a new sensing technique developed at MIT.
A team in MIT’s Quantum Engineering Group has developed a new method to control nanoscale diamond sensors, which are capable of measuring even very weak magnetic fields. The researchers present their work this week in the journal Nature Communications.
The new control technique allows the tiny sensors to monitor how these magnetic fields change over time, such as when neurons in the brain transmit electrical signals to each other. It could also enable researchers to more precisely measure the magnetic fields produced by novel materials such as the metamaterials used to make superlenses and “invisibility cloaks.”
In 2008 a team of researchers from MIT, Harvard University, and other institutions first revealed that nanoscale defects inside diamonds could be used as magnetic sensors.
The naturally occurring defects, known as nitrogen-vacancy (N-V) centers, are sensitive to external magnetic fields, much like compasses, says Paola Cappellaro, the Esther and Harold Edgerton Associate Professor of Nuclear Science and Engineering (NSE) at MIT.
Defects inside diamonds are also known as color centers, Cappellaro says, as they give the gemstones a particular hue: “So if you ever see a nice diamond that is blue or pink, the color is due to the fact that there are defects in the diamond.”
The N-V center defect consists of a nitrogen atom in place of a carbon atom and next to a vacancy — or hollow — within the diamond’s lattice structure. Many such defects within a diamond would give the gemstone a pink color, and when illuminated with light they emit a red light, Cappellaro says.
To develop the new method of controlling these sensors, Cappellaro’s team first probed the diamond with green laser light until they detected a red light being emitted, which told them exactly where the defect was located.
They then applied a microwave field to the nanoscale sensor, to manipulate the electron spin of the N-V center. This alters the intensity of light emitted by the defect, to a degree that depends not only on the microwave field but also on any external magnetic fields present.
To measure external magnetic fields and how they change over time, the researchers targeted the nanoscale sensor with a microwave pulse, which switched the direction of the N-V center’s electron spin, says team member and NSE graduate student Alexandre Cooper. By applying different series of these pulses, acting as filters — each of which switched the direction of the electron spin a different number of times — the team was able to efficiently collect information about the external magnetic field.
They then applied signal-processing techniques to interpret this information and used it to reconstruct the entire magnetic field. “So we can reconstruct the whole dynamics of this external magnetic field, which gives you more information about the underlying phenomena that is creating the magnetic field itself,” Cappellaro says.
The team used a square of diamond three millimeters in diameter as their sample, but it is possible to use sensors that are only tens of nanometers in size. The diamond sensors can be used at room temperature, and since they consist entirely of carbon, they could be injected into living cells without causing them any harm, Cappellaro says.
One possibility would be to grow neurons on top of the diamond sensor, to allow it to measure the magnetic fields created by the “action potential,” or signal, they produce and then transmit to other nerves.
Previously, researchers have used electrodes inside the brain to “poke” a neuron and measure the electric field produced. However, this is a very invasive technique, Cappellaro says. “You don’t know if the neuron is still behaving as it would have if you hadn’t done anything,” she says.
Instead, the diamond sensor could measure the magnetic field noninvasively. “We could have an array of these defect centers to probe different locations on the neuron, and then you would know how the signal propagates from one position to another one in time,” Cappellaro says.
In experiments to demonstrate their sensor, the team used a waveguide as an artificial neuron and applied an external magnetic field. When they placed the diamond sensor on the waveguide, they were able to accurately reconstruct the magnetic field. Mikhail Lukin, a professor of physics at Harvard, says the work demonstrates very nicely the ability to reconstruct time-dependent profiles of weak magnetic fields using a novel magnetic sensor based on quantum manipulation of defects in diamond.
“Someday techniques demonstrated in this work may enable us to do real-time sensing of brain activity and to learn how they work,” says Lukin, who was not involved in this research. “Potential far-reaching implications may include detection and eventual treatment of brain diseases, although much work remains to be done to show if this actually can be done,” he adds. |
|