Texas A&M researchers, along with computer scientists at Intel Labs, have designed a tool to identify the source of errors caused by software updates.
We've all shared the frustration -- software updates that are intended to make our applications run faster inadvertently end up doing just the opposite. These bugs, dubbed in the computer science field as performance regressions, are time-consuming to fix since locating software errors normally requires substantial human intervention.
To overcome this obstacle, researchers at Texas A&M University, in collaboration with computer scientists at Intel Labs, have now developed a complete automated way of identifying the source of errors caused by software updates. Their algorithm, based on a specialized form of machine learning called deep learning, is not only turnkey, but also quick, finding performance bugs in a matter of a few hours instead of days.
"Updating software can sometimes turn on you when errors creep in and cause slowdowns. This problem is even more exaggerated for companies that use large-scale software systems that are continuously evolving," said Dr. Abdullah Muzahid, assistant professor in the Department of Computer Science and Engineering. "We have designed a convenient tool for diagnosing performance regressions that is compatible with a whole range of software and programming languages, expanding its usefulness tremendously."
The researchers described their findings in the 32nd edition of Advances in Neural Information Processing Systems from the proceedings of the Neural Information Processing Systems conference in December.
To pinpoint the source of errors within a software, debuggers often check the status of performance counters within the central processing unit. These counters are lines of code that monitor how the program is being executed on the computer's hardware in the memory, for example. So, when the software runs, counters keep track of the number of times it accesses certain memory locations, the time it stays there and when it exits, among other things. Hence, when the software's behavior goes awry, counters are again used for diagnostics.
"Performance counters give an idea of the execution health of the program," said Muzahid. "So, if some program is not running as it is supposed to, these counters will usually have the telltale sign of anomalous behavior."
However, newer desktops and servers have hundreds of performance counters, making it virtually impossible to keep track of all of their statuses manually and then look for aberrant patterns that are indicative of a performance error. That is where Muzahid's machine learning comes in.
By using deep learning, the researchers were able to monitor data coming from a large number of the counters simultaneously by reducing the size of the data, which is similar to compressing a high-resolution image to a fraction of its original size by changing its format. In the lower dimensional data, their algorithm could then look for patterns that deviate from normal.
When their algorithm was ready, the researchers tested if it could find and diagnose a performance bug in a commercially available data management software used by companies to keep track of their numbers and figures. First, they trained their algorithm to recognize normal counter data by running an older, glitch-free version of the data management software. Next, they ran their algorithm on an updated version of the software with the performance regression. They found that their algorithm located and diagnosed the bug within a few hours. Muzahid said this type of analysis could take a considerable amount of time if done manually.
In addition to diagnosing performance regressions in software, Muzahid noted that their deep learning algorithm has potential uses in other areas of research as well, such as developing the technology needed for autonomous driving.
"The basic idea is once again the same, that is being able to detect an anomalous pattern," said Muzahid. "Self-driving cars must be able to detect whether a car or a human is in front of it and then act accordingly. So, it's again a form of anomaly detection and the good news is that is what our algorithm is already designed to do."
Other contributors to the research include Dr. Mejbah Alam, Dr. Justin Gottschlich, Dr. Nesime Tatbul, Dr. Javier Turek and Dr. Timothy Mattson from Intel Labs.
Amy Halbert | EurekAlert!
Using sound and light to generate ultra-fast data transfer
11.02.2020 | University of Leeds
Project CoPDA: DFKI Laboratory Niedersachsen teaches dynamic knowledge to robots for a better human-machine interaction
06.02.2020 | Deutsches Forschungszentrum für Künstliche Intelligenz GmbH, DFKI
After helping develop a new approach for organic synthesis -- carbon-hydrogen functionalization -- scientists at Emory University are now showing how this approach may apply to drug discovery. Nature Catalysis published their most recent work -- a streamlined process for making a three-dimensional scaffold of keen interest to the pharmaceutical industry.
"Our tools open up whole new chemical space for potential drug targets," says Huw Davies, Emory professor of organic chemistry and senior author of the paper.
Superconductivity approaching room temperature may be possible in hydrogen-rich compounds at much lower pressures than previously expected
Reaching room-temperature superconductivity is one of the biggest dreams in physics. Its discovery would bring a technological revolution by providing...
At the end of December 2019, the first cases of pneumonia caused by a novel coronavirus were reported from the Chinese city of Wuhan. Since then, infections...
A team of researchers from Switzerland, the US and Poland have found evidence of a uniquely high density of hydrogen atoms in a metal hydride. The smaller spacings between the atoms might enable packing significantly more hydrogen into the material to a point where it could begin to superconduct at room temperature and ambient pressure.
The scientists conducted neutron scattering experiments at the Oak Ridge National Laboratory (ORNL) in the US on samples of zirconium vanadium hydride at...
An international research group with Dr. Longjian Xie from the Bavarian Research Institute of Experimental Geochemistry & Geophysics (BGI) of the University of Bayreuth has succeeded for the first time in measuring the viscosity that molten solids exhibit under the pressure and temperature conditions found in the lower earth mantle. The data obtained support the assumption that a bridgmanite-enriched rock layer was formed during the early history of the earth at a depth of around 1,000 kilometres – at the border to the upper mantle.
In addition, the data also provides indications that the lower mantle contains larger reservoirs of materials that originated in an early magma ocean and have...
12.02.2020 | Event News
16.01.2020 | Event News
15.01.2020 | Event News
12.02.2020 | Life Sciences
12.02.2020 | Life Sciences
12.02.2020 | Event News